We rationally designed and developed a group I intron-derived ribozyme that binds an exogenous RNA according to sequence, removes a designated central fragment from the RNA, and ligates the remaining RNA halves back together. We termed this the trans excision-splicing reaction (TES reaction), and the University has patented the technology. This reaction is similar to the self-splicing reaction in that the region to be excised is analogous to the intron itself, but the substrate is exogenous. We characterized this ribozyme in many ways and discovered that the ribozyme could be modified to target a variety of sequences (perhaps in the distant future including therapeutic targets) and that as many as 28 nucleotides and as little as a single nucleotide could be excised.
We also noted that the two-step reaction was inefficient due to the dissociation of one of the intermediates between the reaction steps. In order to overcome this, we conducted a study to identify what molecular recognition components were responsible for keeping this intermediate bound to the ribozyme. We found that this intermediate is bound into two helices (called P9.0 and P10), which was not surprising, but that each helix, although acting distinctly from each other, could be elongated to increase the thermodynamic stability of the bound intermediate complex, thus greatly enhancing the effectiveness of the reaction.
We then studied the sequence requirements of the TES reaction at the two sites of RNA catalysis: the 5’ splice site (which is involved in a base pair) and the 3’ splice site (which is not base paired). Both sites contain nucleosides that are highly conserved throughout group I introns. We determined that the sequence and base pairing requirements for the 5’ splice site were surprisingly lax, which is good in that we can target a wide array of sequences, but which is not good in that this will reduce the sequence specificity of the reaction. We therefore assessed the relative sequence specificity of ribozymes with different base pairs at the 5’ splice site, which will aid in the design of sequence-specific TES ribozymes. The 3’ splice site, however, could not be changed from its conserved guanosine; and so a guanosine has to be the last, or the only, base in the excised region. Note that the two reaction steps of trans excision-splicing mimic the two steps of self-splicing, and the substrate binding components in trans excision-splicing and self-splicing are the same. It is in this way that the new perspective that arises from our studies imparts new knowledge of the self-splicing reaction itself.
We also determined during the course of these studies that at long reaction times, relative to product formation, TES products dissociate and then rebind free ribozyme. This results in degradation of TES products. In the four cases that contain Watson-Crick base pairs (only), however, we found that TES products could rebind ribozyme in the wrong helical register, leading to the formation of cryptic degradation products. Apparently, non-Watson-Crick base pairs at the 5’ splice site aid in the determination of the binding register of reaction substrates. The requirement for site-specific structural perturbation in the helical backbone at the splice junction, then, helps explain why wobble base pairs are present at this position in the self-splicing reaction. In addition, we unexpectedly discovered that the 3’ splice site guanosine (within the substrate) is playing a similar role in defining the proper substrate binding register. This guanosine cannot be playing a comparable role in the self-splicing reaction, however, and so our unique system has in this case helped us discover a new molecular recognition strategy that RNAs can utilize to correctly bind their substrates.