Expanding the Scope of DNA-Templated Polymerisations : New Backbones, Properties and Applications.
Introduction
In biology genetic information is transmitted through the template-directed polymerisations of transcription, translation and replication. After the discovery of the double-helical structure of DNA in 1953,1 the basic elements of the process were readily apparent. However, the fine balance between the thermodynamics of the recognition process of the template and the substrates and the kinetics of the polymerisation that results in the observed fidelity in information transfer has been studied ever since. The high fidelity of the process makes it attractive from a synthetic point of view and as early as 1956 synthetic polymers were explored as templates to direct the course of polymerisation.2 This has achieved some success but was somewhat limited by structurally heterogeneous templates and weak monomer/template association energies. The use of biological macromolecules, which are structurally well defined, helped to solve the heterogeneity problem but the weak association energies difficulty has persisted. In order for catalytic template-directed polymerisation to be successful it must follow a minimal cycle. The cycle involves three steps: 1) Association of substrates to the template; 2) Coupling of substrates to form a new product strand; 3) Dissociation of the product strand from the template to permit another round of synthesis. While catalytic efficiency of the template is represented by the number of turns of the entire cycle per unit time, the accuracy in transmittal of the template information to the product strand (i.e. fidelity) is a function of the selectivity of the molecular recognition step and the kinetics of ligation. Maximizing the difference in relative affinities of the template for the correct versus the incorrect substrate can improve fidelity, but high affinities can also slow the overall catalytic efficiency by restricting product dissociation. An alternative approach to improving product fidelity is to make the ligation of substrates into the product strand a reversible process. Product strand production could then reach a thermodynamic equilibrium that eventually traps the most stable product, which would then correspond to that of the highest fidelity in terms of molecular recognition between the substrates and the template strand. In this way, thermodynamic self-assembly acts in place of the protein machinery that is used by Nature to ensure high fidelity during self-replication. In principle, the ligation step can exploit any reaction that would couple the substrate chain ends. However, the reaction will be constrained by the geometry required to bind the substrates to the template and little information is currently available to place limits on the selection of appropriate reactions. At least with the nucleic acids, the increasing knowledge and synthetic availability of modified backbones may allow the desired association energies and reaction geometry to be rationally engineered into the templates so that these parameters can be optimised.
Previous work in the Lynn group adopted non-phosphodiester-based backbones and polymerising conditions to exploit the reversible ligation scheme described above. By replacing phosphodiester
formation with reductive amination, utilizing pre-equilibrium imine formation on the template prior to imine reduction (i.e. amination), association thermodynamics drove the fidelity of substrate condensation on the template. The results of this reaction, in comparison to phosphodiester condensation, are truly remarkable. For example, the DNA octamer (dAp)8 template catalyses a quantitative conversion of the thymidine monomer 5’-H2N-dT-3’-CH2CHO (T) to the octameric polyamine product, upon reduction of the imine linkages with NaCNBH3. Dimer and tetramer products appear only as intermediates, as can be seen in the HPLC trace. Therefore, this process achieves the first chain length specific reading of a DNA template ever achieved via monomer polymerisation without the aid of protein enzymes. The reaction follows classical step-growth kinetics, distinctly different from that reported with phosphodiester formation. In the absence of the template no reaction occurs. Further experiments have shown that the DNA sequence is read stereo-, sequence and chain-length (up to 32 mer) selectively.
Proposed Studies
My current work focuses on expanding the scope and range of the ligation chemistries that can be utilised in these template-directed polymerisations. The accurate synthetic translation of information encoded in biological macromolecules, exploiting more standard polymerisation reactions, should enable the preparation of a diversity of monodisperse, sequence-specific functional materials which could have a range of potential applications. A number of different strategies are proposed utilising some recent advances in synthetic methodology.
