Our current understanding of biology suggests that life begin with an earlier RNA-based life system (the 'RNA world'), in which RNA played the role of information storage (genes) and catalysis (ribozymes). The search for an artificial replicase ribozyme, one of the essential molecules for the 'RNA world', has involved one decade of in vitro evolution studies. Currently, the best replicase-like ribozymes are RNA polymerase ribozymes that are capable of polymerizing a primer-template up to 14-21 nucleotides. This extension corresponds in an absolute sense to about 1/10 of the extension required to replicate the RNA catalysts. We poorly understand how their structure contributes to polymerization.

To improve the catalytic rate and to better understand the structural basis for extension, we have developed a range of RNA polymerase ribozyme constructs by deleting or rearranging the secondary structural elements of a RNA polymerase ribozyme called B6.61. Our most striking result demonstrates that B6.61 can be separated into two separate molecules ('catalytic domain part' and 'accessory domain part'), which work cooperatively in trans albeit with a 100-fold lower catalytic rate. Interestingly, the activity of this bi-molecular ribozyme is nearly completely recovered by hybridizing the two domains back together, suggesting that the two domains can interact in a flexible way. Other constructs were developed that tether the primer to the ribozyme so as to increase substrate-ribozyme local concentrations. Tethering was achieved by hybridizing a DNA sequence, which is covalently linked via a flexible PEG linker to the 5' end of a primer, to a RNA tag sequence inserted into different spots of the B6.61 ribozyme. We find certain tethering sites that increase the rate of primer extension by as much as 10-fold, whereas other sites that dramatically decrease extension activity. These findings facilitate our understanding of this ribozyme.

Now, we continue to map the contacts between the two major structural domains of this ribozyme, mainly by UV crosslinking. By incorporating 4-thio uridine residues (4sU) into either domain, a range of crosslinks between the 4sU and their spatially closely located residues on the other domain will form by UV irradiation. Several crosslinked residues have been successfully identified and may prove to be functional. In the near future, we hope a model of the RNA polymerase can be built that includes all the structural and functional information gathered in this study.