Abstract
Purpose :
Computational structural modelling is used to identify optimal mRNA target sites for Post-Transcriptional Gene Silencing (PTGS) agents such as shRNAs and ribozymes. Optimal sites are generally single-stranded and accessible in a stable structure in vivo. This approach streamlines later screening and optimization. Experimental steps focus on mature mRNA sequence transcribed from cDNA, without consideration of pre-mRNA structure, wherein the presence of an intron and history of splicing could modify target site accessibility. This study was conducted to determine predicted exon structural differences in Human RHO mature and pre-mRNA, and to analyze effect on target sites of PTGS agents that have been developed in our lab and by other groups as potential treatments for Autosomal Dominant Retinitis Pigmentosa caused by Rhodopsin mutations.
Methods :
Human RHO pre-mRNA (5470 nt) and mature mRNA (1532 nt) sequences were folded at the secondary level using SFold, MFold, and RNAStructure. These algorithms predict structure based on single stranded probability by Boltzmann weighting, minimal folding energy, and local free energy, respectively. The product of these vectors determines the Multiparameter Prediction of RNA accessibility (MPPRNA) and were graphed. Area under the curve (AUC) was calculated at annealing sites for 12 RHO targeting PTGS agents as a measure of accessibility.
Results :
RHO exon 1, 2, and 5 pre- vs. mature mRNA MPPRNA values were significantly different (p<0.05) by one way ANOVA. Target site AUC was significantly different (p<0.05) by 2-way independent t-tests for 7 PTGS agents. Generally, PTGS agents that have targets in exons 2-4 (n=6) were more susceptible to alterations in accessibility than those in exon 1 and 5 (n=6.)
Conclusions :
The history of intron splicing could potentially affect mature mRNA target structure and accessibility for PTGS agents. As most PTGS studies use cDNAs to express mature mRNAs, these may not simulate the optimal needs of PTGS gene therapy on real in vivo processes during a clinical trial (i.e. transcription of target pre-mRNAs from nuclear genes followed by splicing, capping, polyadenylation, and nuclear export for translation.) If the history of splicing imparts accessibility changes in mature target sites, then cDNA experimental outcomes may not be maximally predictive of in vivo performance.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.