Combining single-molecule manipulation and peptide nucleic acid binding studies for unraveling how RNA structures regulate ribosomal frameshifting and alternative splicing
Date of Issue2019-06-10
School of Physical and Mathematical Sciences
RNA structures are involved in regulating many biological activities such as programmed ribosomal frameshifting and pre-mRNA alternative splicing. Studying the structural features, mechanical stabilities, unfolding and folding dynamics of native RNAs and their mutants, as well as the interaction with ligands such as peptide nucleic acids can help facilitate a deep and thorough understanding of the RNAs and the related biological functions. Here, we combined single-molecule manipulation using optical tweezers and peptide nucleic acids binding studies to unravel the biophysical properties of model RNA hairpins and disease-related RNA structures, and their correlations with minus-one ribosomal frameshifting and pre-mRNA alternative splicing activities. Minus-one ribosomal frameshifting produces multiple functional proteins from one viral mRNA through ribosomal backward slippage by one nucleotide. The alternative splicing of exons in a pre-mRNA by spliceosome produces different protein isoforms through the inclusion or exclusion of exons. Project 1. A wobble A∙C pair can be protonated at near physiological pH to form a more stable wobble A+∙C pair. Here, we constructed an RNA hairpin (rHP) and three mutants with one A-U base pair substituted with an A∙C mismatch on the top (near the loop, U22C), middle (U25C) and bottom (U29C) positions of the stem, respectively. Our results on single-molecule mechanical (un)folding using optical tweezers reveal the destabilization effect of A-U to A∙C pair substitution, and protonation-dependent enhancement of mechanical stability facilitated through an increased folding rate, or decreased unfolding rate, or both. Our data show that protonation may occur rapidly upon the formation of apparent mechanical folding transition state. Furthermore, we measured the bulk −1 ribosomal frameshifting efficiencies of the hairpins by a cell-free translation assay. For the mRNA hairpins studied, −1 frameshifting efficiency correlates with mechanical unfolding force at equilibrium and folding rate at around 15 pN. U29C has a frameshifting efficiency similar to that of rHP (~2%). Accordingly, the bottom 2-4 base pairs of U29C may not form under a stretching force at pH 7.3, which is consistent with the fact that the bottom base pairs of the hairpins may be disrupted by ribosome at the slippery site. U22C and U25C have a similar frameshifting efficiency (~1%), indicating that both unfolding and folding rates of an mRNA hairpin in a crowded environment may affect frameshifting. Our data indicate that mechanical (un)folding of RNA hairpins may mimic how mRNAs unfold and fold in the presence of translating ribosomes. Project 2. Base triples in RNA pseudoknots are crucial in maintaining RNA structure, stability, and biological functions. Minor-groove base triples formed between stem 1 and loop 2 of SRV-1 mRNA frameshifting pseudoknot are essential in determining the thermodynamics and mechanical properties of the pseudoknot as well as stimulating the minus-one ribosomal frameshifting. How tertiary base triple formation affects the stability of stem 1 and stem 2 and thus ribosomal frameshifting efficiency is not well understood. We designed a PNA that is expected to be able to invade stem 1 of the SRV-1 pseudoknot to mimic stem 1 unwinding process by a translating ribosome. In addition, we designed a PNA for invading stem 2 in SRV-1 pseudoknot. Our non-denaturing polyacrylamide gel electrophoresis data revealed that a single mutation in loop 2 disrupting the base triple formation in SRV-1 pseudoknot results in tighter binding by both PNAs. Thus, tertiary stem 1-loop 2 base triple interactions in SRV-1 pseudoknot can stabilize the secondary structural components stem 1 and stem 2. We observed that invasion efficiencies of both PNAs positively correlated with the mechanical stability and the frameshifting efficiency of the SRV-1 pseudoknot, and inversely correlated with their unfolding rates. Our work suggests that both stem 1 and stem 2 stabilities in a pseudoknot structure are important for stimulating ribosomal frameshifting. Project 3. Frontotemporal dementia with Parkinsonism linked to chromosome-17 (FTDP-17) is caused by mutations in the microtubule-associated protein tau (MAPT) gene. Alternative splicing of the MAPT cassette exon 10 produces tau isoforms with 4 microtubule-binding repeat domains (4R) upon exon inclusion, or 3 repeats (3R) upon exon skipping. In human neurons, deviations from the physiological 4R:3R ratio of ~1:1 lead to FTDP-17. Certain FTDP-17-associated mutations affect a regulatory hairpin that sequesters the exon 10 5′ splice site (5′ss). These mutations tend to increase the 4R:3R ratio by destabilizing the hairpin, thereby improving 5′ss recognition by U1 small nuclear ribonucleoprotein. Interestingly, a single C-to-G mutation at the 19th nucleotide in intron 10 (C19G or +19G) reduces exon 10 inclusion significantly from 56% to 1%, despite the disruption of a G-C base pair in the bottom stem of the hairpin. Here, we show by biophysical characterization including thermal melting and single-molecule mechanical unfolding using optical tweezers, that the +19G mutation alters the structure of the bottom stem, resulting in the formation of a new bottom stem with enhanced stability. Some of the disease-causing point mutations in the top stem may also result in local structural rearrangement within the top stem. We exploited the effect of the +19G mutation on the known FTDP-17-linked hairpin top stem-destabilizing mutants. The cell culture alternative splicing patterns of a series of minigenes containing the disease-causing mutations and artificial mutations reveal that the splicing activities of the mutants with destabilizing mutations on the top stem can be compensated in position-dependent manner by a +19G mutation in the bottom stem. A local 5-base pair stem surrounding the 5′ss site (involving residues −2 to +3 and +12 to +15) is critical in determining the alternative splicing pattern, and a point mutation within this local stem causes an increased exon 10 inclusion from 56% to above 95%, with the splicing level not compensated by a stabilizing mutation such as +19G in the bottom stem. We observed an excellent correlation between exon 10 inclusion level and the mechanical unfolding rate at 10 pN, consistent with the fact that the unfolding of the splice site hairpins is aided by helicases. Thus, mechanical unfolding directly reveals the structural changes upon single-nucleotide mutations and can mimic how the splice site hairpins are unfolded in the cell for the recognition by U1 snRNA.