Understanding the Structural Properties of Triple Helices and a Triple-Stranded RNA-Binding Protein

Though triple helices were deduced to form in a test tube over sixty years ago, the function of triple helices in nature is only beginning to be appreciated. My graduate work focused on pyrimidine-motif triple helices, whereby a pyrimidine-rich third strand binds in the parallel orientation along the purine-rich strand in the major groove of a double helix. Herein, I explored two questions about triple helices: (i) which base triples stabilize RNA•DNA-DNA triple helices? and (ii) how does a protein recognize an RNA triple helix?

Noncoding RNAs are hypothesized to bind to genomic DNA, forming an RNA•DNA-DNA triple helix, to regulate gene expression. However, beyond the canonical U•A-T and C•G-C base triples, the stability of base triples that compose RNA•DNA-DNA triple helices is unknown. Therefore, I designed a 22-base triple U•A-T-rich RNA•DNA-DNA triple helix to systematically determine the stability of the triple helix when a single central base triple, Z•X-Y (where Z = C, U, A, G, and X-Y = A-T, G-C, T-A, C-G), is varied, using an electrophoretic mobility shift assay to examine the binding between the RNA and double-stranded DNA. Results indicate that the canonical U•A-T and C•G-C base triples are among the most stable, seven non-canonical base triples have stabilities within two-fold of the canonical U•A-T base triple, and multiple consecutive non-canonical base triples completely disrupted the formation of the RNA•DNA-DNA triple helix. Additionally, an RNA shorter than 19-nucleotides disrupts triple helix formation, though longer triple helices did not lead to tighter binding. Further examination of the RNA•DNA-DNA triple helix revealed that natural RNA modifications either have no observable effect or destabilize the triple helix. These studies have led to a better understanding of RNA•DNA-DNA base triples that may form in nature.

To understand how a protein interacts with a triple-stranded RNA, I utilized the methyltransferase-like protein 16 (METTL16), which binds to the triple helix from the metastasis-associated lung adenocarcinoma transcript 1 (MALAT1) long noncoding RNA. Using a variety of biochemical assays, I show that the C-G doublet that interrupts the triple helix and its location within the triple helix is necessary for binding to METTL16, though the identity of the doublet can be any of the four Watson-Crick base pairs. Further, both the N- and C-terminal domains of METTL16 bind to the MALAT1 triple helix, with the C-terminal domain providing the specificity for the C-G doublet while the N-terminal interacts with structural features below the triple helix. Overall, this work has increased our knowledge of base triples that stabilize triple helices and of protein-triple helix interactions