The effects of structure on pre-mRNA processing and stability

Abstract

Pre-mRNA molecules can form a variety of structures, and both secondary and tertiary structures have important effects on processing, function and stability of these molecules. The prediction of RNA secondary structure is a challenging problem and various algorithms that use minimum free energy, maximum expected accuracy and comparative evolutionary based methods have been developed to predict secondary structures. However, these tools are not perfect, and this remains an active area of research. The secondary structure of pre-mRNA molecules can have an enhancing or inhibitory effect on pre-mRNA splicing. An example of enhancing structure can be found in a novel class of introns in zebrafish. About 10% of zebrafish genes contain a structured intron that forms a bridging hairpin that enforces correct splice site pairing. Negative examples of splicing include local structures around splice sites that decrease splicing efficiency and potentially cause mis-splicing leading to disease. Splicing mutations are a frequent cause of hereditary disease. The transcripts of disease genes are significantly more structured around the splice sites, and point mutations that increase the local structure often cause splicing disruptions. Post-splicing, RNA secondary structure can also affect the stability of the spliced intron and regulatory RNA interference pathway intermediates, such as pre-microRNAs. Additionally, RNA secondary structure has important roles in the innate immune defense against viruses. Finally, tertiary structure can also play a large role in pre-mRNA splicing. One example is the G-quadruplex structure, which, similar to secondary structure, can either enhance or inhibit splicing through mechanisms such as creating or obscuring RNA binding protein sites.

Keywords: Disease; ESE; ESS; G quadruplex; RNA processing; Secondary structure; Simple repeats; Splice site; Splicing; Zebrafish.

Figure 1. Detecting Local Stable Structures

RNA is examined for local stable structures using a sliding window approach. Within window size of W only base pairings within a maximum distance apart (L) are considered. The orange arcs represent possible valid base pairings.

Figure 2. Complementary dinucleotide repeats at the ends of Zebrafish introns drive secondary structure

A. Minimum fold energy of Zebrafish (AC)m-(GT)n introns (red) and all Zebrafish introns (blue). Introns are binned by length and minimum fold energy is calculated using RNAfold. B. All possible dinucleotide repeats are added in silico to Zebrafish introns 20 nucleotides downstream of the 5’ss, and the complementary dinucleotide repeat is added 20 nucleotides upstream of the 3’ss. Synthetic introns are folded using RNAfold and the number of introns with predicted repeats base-pairing to form a hairpin structure is counted.

Figure 3. Structure-mediate pre-mRNA splicing

A. In zebrafish, complementary dinucleotide repeats at two ends of the intron bring splice sites to close proximity and facilitate splicing. The structural effect could overcome the absence of the essential splicing factor U2AF2. B. Complementary G/C triplet repeats stabilize human introns structure potentially enhancing splicing similar to A. C. In silico experiments suggest that repeats of complementary dinucleotides can bridge intron ends in the predicted structure of some loci. D. Introns flanking hyper-polymorphic exons are stabilized by secondary structure to ensure splicing.

Figure 4. Minimum free energy (ΔG) of disease gene splice sites

A. RNAfold minimum free energy score (ΔG) of 3’ss in HGMD disease genes (HGMD) and the remaining genes in the genome (non-HGMD). B. RNAfold minimum free energy score (ΔG) of 5’ss in HGMD disease genes (HGMD) and the remaining genes in the genome (non-HGMD).

Figure 5. Role of RNA secondary structures in human diseases

Structural accessibilities around splice sites are important for splicing. Degree of openness (predicted %unpaired bases) in the 3’ splice sites (A) and 5’ splice sites (B) correlates with increased splicing efficiency. (C) Thermodynamically less stable transcripts are more susceptible for exonic splicing mutations (ESM). (D) T to G mutation (blue and red highlight, respectively) in HMBS gene resulted in more stable structure. (E) Mutant splicing relative to wildtype in some mutations that stabilized the RNA structure in the different stages of the spliceosome assembly.

Figure 6. Structure models of the stable LAT intron lariat intermediate and predicted free energies

(A) Lariat model structure of the LAT intron from the 5' splice site to the branchpoint nucleotide (loop of lariat). (B) Lariat model structure of the LAT intron from the branch point nucleotide to the 3' splice site (tail of lariat).

Figure 7. Structure of a G-Quadruplex

A G-Quadruplex is composed of 4 interspersed G-triplets which fold into stacked G-quartets (gray diamonds). Each guanine residue in a G-quartet is associated with its neighboring guanines through Hoogsteen base-pairing.

Figure 8. Alternative splicing profile for G4 stabilizing ligands

A. Number of alternative splicing events. SE: skipped exon, A5SS: alternative 5’ splice site, A3SS: alternative 3’ splice site, MXE: mutually-exclusive exons, RI: retained intron. B. Percent of SE events in which there are four or more G-triplets in at least one flanking intron.