Challenges and opportunities in technologies and methods for lncRNA structure determination

Abstract

Long non-coding RNAs (lncRNAs) play pivotal roles in diverse cellular processes ranging from gene regulation and chromatin remodeling to RNA stability and epigenetic modifications. Despite the identification of approximately 95,000 lncRNA genes in humans, our understanding of their structure–function relationships remains very limited. This review examines the current state of lncRNA structure determination. We briefly discuss the advantages and limitations of experimental approaches—including chemical probing methods such as SHAPE and DMS—as well as the challenges inherent to computational predictions, particularly given RNA's dynamic nature, structural heterogeneity, and the energy degeneracy of its building blocks. The review also highlights the difficulties in predicting long-range interactions, including pseudoknots, which are essential for global folding of large RNAs, and discusses how elevated, nonphysiologically Mg2⁺ concentrations used in many experiments can distort our perception of native RNA conformations. Recent advances in cryo-electron microscopy and atomic force microscopy, coupled with machine learning algorithms, offer promising strategies to capture the realistic conformational landscapes of RNAs, including lncRNAs, under near-physiological conditions. These advances have the potential to redefine our understanding of lncRNA architectures, their structural dynamics, and how they influence cellular functions, ultimately informing future directions of lncRNA research and opening new frontiers such as structure-based drug discovery and therapeutic interventions targeting lncRNAs.