Current Organic Chemistry - Volume 6, Issue 14, 2002

Small, zwitterionic peptides (2-6 residues) are flexible molecules that occur in solution as an ensemble of conformational forms in dynamic equilibrium. They occur widely in Nature as components of the peptide pool produced by protein hydrolysis. Molecular modelling provides information on the conformations they adopt in aqueous solution. Contrary to expectations, they exist in a limited range of conformational forms that arise from particular combinations of backbone torsions. From the computed energies of the conformers, those with similar conformations can be aggregated and a profile of the conformer population displayed using a 3D pseudo- Ramachandran plot. Several different generic peptide transporters have evolved with complementary substrate specificities in response to these varied conformational forms. Molecular recognition templates (MRTs) define the stereo-electronic features of particular conformations that determine peptide recognition and binding by these transporters. The binding and transport of any peptide correlates with the proportion of its conformers that occur in a particular MRT. Specific conformer binding appears generally to involve little or no ligand rearrangement, which minimises conformational entropy loss. Conformational analysis of peptide analogues, evaluation of their similarities to normal peptides and comparison of their conformer profiles with MRTs, allows estimates of their molecular recognition as putative bioactive ligands for such proteins, this is illustrated for a variety of pseudopeptides, isosteres and conformationally-constrained analogues. For peptide therapeutics, such as β-lactam antibiotics, inhibitors of angiotensin-converting enzyme, antiviral and chemotherapeutic agents, this approach can be used to evaluate in silico virtual libraries of drug analogues to estimate their potential for targeted delivery and oral availability.

Oxidation of retinol to retinaldehyde is an important first step in the biosynthesis of a biologically active retinoic acid, which serves as an activating ligand for retinoic acid receptors. Retinol can be oxidized by a variety of different oxidoreductases found in both the cytosolic and the microsomal fractions of the cells. In the cytosol, the best characterized retinol-active dehydrogenases are the dimeric 40 kDa-subunit zinc metalloenzymes that belong to the superfamily of medium-chain alcohol dehydrogenases. Microsomes contain retinol-active dehydrogenases with 35 kDa-subunit molecular weight that belong to the short-chain dehydrogenase / reductase superfamily. Both types of enzymes recognize other physiological substrates besides retinol and exist in multiple isozymic forms. Isozymes exhibit different tissue distribution and have various catalytic efficiencies for oxidation of retinol. The relative physiological importance of the multiple forms of retinol dehydrogenases is currently being debated. This review summarizes some of our data on the identification and characterization of the candidate cytosolic and microsomal retinol dehydrogenases.

G protein-coupled receptors (GPCRs) are heptahelical transmembrane proteins. They transduce a large variety of extracellular signals, allowing perception of taste, odor, hormones, and light. Binding of an extracellular ligand induces structural changes in the cytosolic domain of a GPCR which, thereby, catalyzes nucleotide exchange in a G protein. Rhodopsins, the visual pigments, are prototypical GPCRs that are activated by photoisomerization of covalently bound 11-cis retinal. Unlike other GPCRs, bovine rhodopsin and transducin, its cognate G protein, can be prepared from cow eyes in large quantities for spectroscopic and biochemical investigations. Rhodopsin is the best studied GPCR and the only one for which an X-ray structure has been solved. Structural information together with the wealth of biophysical data on native and recombinant rhodopsins allows to determine structure function relationships that are relevant to GPCR-dependent signaling in general. Here, results from Fourier-transforminfrared (FTIR) spectrocopic studies of rhodopsin and measurements of nucleotide-dependent transducin fluorescence are reviewed. Intra- and intermolecular processes during signaling by the photoreceptor have thus been identified and analyzed kinetically. Recent applications of these techniques concern rhodopsin transducin coupling in synthetic lipidic matrices and analysis of drug action at the receptor G protein interface.The data are discussed in the context of the crystal structure of rhodopsin and additional biochemical information if required for the understanding of the spectroscopic results.

There has been a great deal of interest in recent years in using boronic acid as the recognition motif for the development of sensors. Because boronic acids can form tight and reversible complexes with diol compounds such as carbohydrates, the majority of the efforts, led by the Shinkai group, have been on the development of sensors for carbohydrates. Boronic acids are also known to selectively recognize fluoride among halides and other anions. Therefore, there have also been efforts in using boronic acid compounds for the development fluoride sensors. This paper reviews the progress in this field during the last five years.

Computational methods available for the calculation of relative and absolute binding affinities (free energy simulations, continuum electrostatics, linear interaction energy approximations, and empirical solvation models) are reviewed together with recent applications to biological systems. The decomposability of the binding free energy into physically meaningful components is examined and results obtained for these components are presented. Some of these components, such as the direct interactions, the translational / rotational entropy loss, and the desolvation free energy are well recognized. Recent calculations have shown that the translational / rotational entropy loss is not as large as some theoretical calculations have previously suggested because of substantial residual movements in the bound complex. Recent work also points to the importance of contributions that are often neglected in binding affinity calculations, such as the protein reorganization energy and, for flexible ligands, the ligand reorganization energy. Future work should concentrate on the improvement of the energy functions and simulation protocols for the achievement of more precise and accurate predictions.

In the last years DNA triplexes have been the subject of a very intense research effort, which has been summarized in a vast number of books and reviews (see for example refs. [1-16]). After a general introduction to DNA polymorphism and the structure of triple helices, we will examine here general approaches to the stabilization of triple helices. Methods based in external binding to DNA and backbone modification will be briefly summarized, and the reader will be addressed to more specific reviews. Finally, we will survey strategies for triplex stabilization based on modification of nucleobases, paying particular attention to recent advances in the field.