Computational identification of potential tandem pseudoknots in the 5'UTR and start codon regions of human mRNAs.

Abstract

Tandem pseudoknots are distinctive RNA structures with two or more pseudoknots arranged in close proximity. While known in viral and bacterial RNAs, their presence in eukaryotic systems, especially human mRNAs, remains unexplored. Using computational tools, we conducted a transcriptome-scale study to identify potential pseudoknots in human mRNAs. Analysis of positional data revealed numerous potential tandem pseudoknots, of which we present 50 representative cases found in the 5' UTR and start codon regions. These arrays contain two or three pseudoknots, typically without intervening sequences, allowing coaxial stacking of stems to form quasi-continuous helices averaging 30 base pairs. The stem regions are GC-rich, with stabilities comparable to or exceeding known functional pseudoknots. Modeling studies of three selected tandem pseudoknots demonstrated their stereochemical feasibility and revealed both common features and diversities. These potential structures are distributed across the 5' UTR and start codon regions, often spanning the entire 5' UTR and extending nearly 100 nucleotides into the coding region. Most mRNAs containing these tandem pseudoknots encode proteins with essential functions implicated in various diseases, with many serving as drug targets. These findings suggest potential prevalence of tandem pseudoknots in regulating mRNA functions. Further investigation and characterization of these structures would enhance our understanding of RNA biology and reveal new regulatory mechanisms that could be leveraged for therapeutic benefits.