Secondary Structural Model of Human MALAT1 Reveals Multiple Structure-Function Relationships.

Phillip J Mccown,Matthew C Wang,Luc Jaeger,Jessica A Brown

doi:10.3390/ijms20225610

Abstract

Human metastasis-associated lung adenocarcinoma transcript 1 (MALAT1) is an abundant nuclear-localized long noncoding RNA (lncRNA) that has significant roles in cancer. While the interacting partners and evolutionary sequence conservation of MALAT1 have been examined, much of the structure of MALAT1 is unknown. Here, we propose a hypothetical secondary structural model for 8425 nucleotides of human MALAT1 using three experimental datasets that probed RNA structures in vitro and in various human cell lines. Our model indicates that approximately half of human MALAT1 is structured, forming 194 helices, 13 pseudoknots, five structured tetraloops, nine structured internal loops, and 13 intramolecular long-range interactions that give rise to several multiway junctions. Evolutionary conservation and covariation analyses support 153 of 194 helices in 51 mammalian MALAT1 homologs and 42 of 194 helices in 53 vertebrate MALAT1 homologs, thereby identifying an evolutionarily conserved core that likely has important functional roles in mammals and vertebrates. Data mining revealed that RNA modifications, somatic cancer-associated mutations, and single-nucleotide polymorphisms may induce structural rearrangements that sequester or expose binding sites for several cancer-associated microRNAs. Our findings reveal new mechanistic leads into the roles of MALAT1 by identifying several intriguing structure–function relationships in which the dynamic structure of MALAT1 underlies its biological functions.

Highlights

Over 15,000 long noncoding RNAs have been discovered in humans, yet little is known about their structure and function [1,2]
Metastasis-associated lung adenocarcinoma transcript 1 (MALAT1), known as nuclear-enriched abundant transcript 2 (NEAT2), is an 8708-nucleotide long noncoding RNA (lncRNA) that is conserved in vertebrates and localizes to nuclear speckles [3,4]
MALAT1 reads were extracted from these datasets and used to model the secondary structure of four known structures in MALAT1: the hnRNPC- and hnRNPG-binding hairpins, the triple helix-containing hairpin and the tRNA-like structure of MALAT1-associated small cytoplasmic RNA (mascRNA) (Figure S1) [11,12,17,18]

Summary

Introduction

Over 15,000 long noncoding RNAs (lncRNAs) have been discovered in humans, yet little is known about their structure and function [1,2]. Increased MALAT1 expression leads to sponging of miR-217 in pancreatic cancer cells, increasing oncogenic KRAS expression [16] These interactions with proteins, U1 snRNA, miRNAs, and other factors are likely modulated in part by the MALAT1 structure. At the 3 end of MALAT1, a triple helix protects mature MALAT1 from a rapid nuclear RNA decay pathway, while a tRNA-like structure, known as MALAT1-associated small cytoplasmic RNA (mascRNA), facilitates 3 -end processing by RNase P [17,18] Both the triple helix and mascRNA are highly conserved structures that enabled the discovery of unannotated MALAT1 homologs and MALAT1-like lncRNAs [4,19,20]

Methods

Results

Conclusion