Molecular design modelling and rationalisation of mode of action of synthetic retinoids in cellular development processes.

Stem cell biology has potential to provide cells and tissues of specific types for use in research, the development of pharmaceuticals and cell replacement therapies. Engineering specific tissue types from stem cells is dependent on being able to direct cell differentiation in a repeatable manner and are stable. All trans-retinoic acid (ATRA) is used to induce cell differentiation in vitro, however, it degrades resulting in products that can effect differentiation in ways not intended. To improve the efficiency of cell differentiation control, we have produced and tested a number of synthetic retinoids which have significantly improved chemical and physical stability. These compounds effect developmental change in human pluripotent stem cells and neuroblastoma cells, and like their natural counterparts, they induce a number of effects, including the formation of mature cells types, including neurons, and cell apoptosis. We have begun to build up a body of results, the first examples of which are in press, and we now need to be able to understand structure versus biological effect in order to design the next generations of more-specific small molecules. Of particular importance is the ability to design compounds which do not just act as analogues of either ATRA or the other RA isomers, but which are benign in terms of the building blocks used and do not elicit toxic effects if used for future tissue-regeneration applications. We propose to carry out molecular modelling studies to examine structure-activity relationships of synthetic retinoids compared to natural systems. Sufficient biological data are available, based on over 30 synthetic structures and 3 natural RA compounds. The competitive binding properties of many of these compounds to nuclear receptors are known, with further compounds being evaluated regularly. This provides the data required for a CASE project which will develop a molecular-level understanding of the mode of action of each of the synthetic compounds, from which new series of compounds will be designed. ATRA is a relatively flexible ligand which binds with high affinity to all RARs. Similarly, the relative flexibility of 9-cis-RA allows it to bind to RXRs as well as to RARs which results in low specificity of receptor targeting in therapeutic applications, giving side effects that limit the therapeutic potential of existing retinoids. Synthetic ligands can be designed to be conformationally rigid, allowing considerable scope for targeting specific RARs with higher specificity. The structure of RARs, particularly the ligand binding pocket of RARgamma and its interactions with a range of natural and synthetic ligands, has been elucidated at high resolution. These data permit the ligand-binding domains of RARalpha and RARbeta and their interactions with RARalpha and RARbeta-specific agonists to be modelled, facilitating the design of novel compounds with subtype specificity. This would be achieved by modelling the interactions between the ligands and the binding domain of different RARs, particularly with respect to the conformation of the carboxy-terminal helix 12 which caps the binding pocket on the receptor and the conformation of which is essential for co-activator/co-repressor recruitment and dissociation. The project will involve: a) all compounds will be modelled using molecular mechanics, molecular dynamics, DFT and/or MP2 calculations, and compared with each other, and experimental structures determined by single-crystal X-ray diffraction; b) each structure will subjected to virtual binding studies with RAR and RXR systems, to obtain an approximate rank order of estimated binding capability; c) these data will be compared with actual SAR screening data to look for correlations between virtual and actual binding. Second generation systems will be designed for further synthesis and screening.