Dynamic molecular networks: from synthetic receptors to self-replicators

Dynamic combinatorial libraries (DCLs) are molecular networks in which the network members exchange building blocks. The resulting product distribution is initially under thermodynamic control. Addition of a guest or template molecule tends to shift the equilibrium towards compounds that are receptors for the guest. This Account gives an overview of our work in this area. We have demonstrated the template-induced amplification of synthetic receptors, which has given rise to several high-affinity binders for cationic and anionic guests in highly competitive aqueous solution. The dynamic combinatorial approach allows for the identification of new receptors unlikely to be obtained through rational design. Receptor discovery is possible and more efficient in larger libraries. The dynamic combinatorial approach has the attractive characteristic of revealing interesting structures, such as catenanes, even when they are not specifically targeted. Using a transition-state analogue as a guest we can identify receptors with catalytic activity. Although DCLs were initially used with the reductionistic view of identifying new synthetic receptors or catalysts, it is becoming increasingly apparent that DCLs are also of interest in their own right. We performed detailed computational studies of the effect of templates on the product distributions of DCLs using DCLSim software. Template effects can be rationalized by considering the entire network: the system tends to maximize global host-guest binding energy. A data-fitting analysis of the response of the global position of the DCLs to the addition of the template using DCLFit software allowed us to disentangle individual host-guest binding constants. This powerful procedure eliminates the need for isolation and purification of the various individual receptors. Furthermore, local network binding events tend to propagate through the entire network and may be harnessed for transmitting and processing of information. We demonstrated this possibility in silico through a simple dynamic molecular network that can perform AND logic with input and output in the form of molecules. Not only are dynamic molecular networks responsive to externally added templates, but they also adjust to internal template effects, giving rise to self-replication. Recently we have started to explore scenarios where library members recognize copies of themselves, resulting in a self-assembly process that drives the synthesis of the very molecules that self-assemble. We have developed a system that shows unprecedented mechanosensitive self-replication behavior: depending on whether the solution is shaken, stirred or not agitated, we have obtained a hexameric replicator, a heptameric replicator or no replication, respectively. We rationalize this behavior through a mechanism in which replication is promoted by mechanically-induced fragmentation of self-assembled replicator fibers. These results represent a new mode of self-replication in which mechanical energy liberates replicators from a self-inhibited state. These systems may also be viewed as self-synthesizing, self-assembling materials. These materials can be captured photochemically, converting a free-flowing fiber solution into a hydrogel through photo-induced homolytic disulfide exchange.