| Riboswitches are regulatory mRNA elements most commonly found in the 5'' leader region of bacterial mRNA transcripts. They are able to regulate expression of the downstream gene product by virtue of their ability to form one of two mutually exclusive secondary structures in a ligand dependent manner. Riboswitches recognize a diverse array of ligands, ranging in complexity from magnesium ions to amino acids and nucleic acid precursors to whole tRNA molecules. The aptamer domain of the riboswitch acts a receptor to bind the target ligand with extremely high specificity as to avoid responding to ligand analogs subtly different from the cognate ligand. The mechanism by which the binding event is transduced into a change in gene expression is thought to involve stabilization of a metastable aptamer domain for a given amount of time. The element of time is critical because riboswitches have evolved under the kinetic constraints inherent to the process of transcription. That is, ligand must bind and stabilize the aptamer domain before the terminator stem in the expression platform is fully transcribed. Because the act of ligand binding is at the heart of the riboswitch mechanism, it is crucial to understand how ligand binding specificity is achieved and how ligand binding stabilizes the aptamer domain.;To further define the rules governing ligand binding specificity determinants for riboswitches in general, I have shown how the S-adenosylhomocysteine (SAH) riboswitch achieves a >1,000-fold specificity for SAH over the closely related compound S-adenosylmethionine, and I have shown how ligand specificity can be swapped for the purine riboswitch using a rational, structure-based mutagenesis approach. Through these studies I have also shown how SAH binding directly stabilizes the aptamer domain, providing further evidence for a generalized regulatory mechanism by riboswitches. Furthermore, through structure-based mutagenesis I have shown that only one nucleotide substitution in the purine riboswitch is required to switch ligand specificity from favoring guanine to 2'deoxyguanosine. This result provides evidence for a model of molecular evolution dating back to the 1960s, which postulates a sequence landscape in which every functional RNA (or protein) can be accessed by starting with one sequence on the grid and ending with another by making a series of mutations that do not abrogate activity. Finally, I have undertaken an investigation into the evolutionary relationship between three known classes of SAM riboswitches. Through this ongoing research, I hope to highlight the modular nature of tertiary architecture and its role in stabilizing a biologically functional ligand binding core. |
