SHAPE Probing Reveals Human rRNAs Are Largely Unfolded in Solution

Area of Science:

• Molecular Biology
• Structural Biology
• Genomics

Background:

• Accurate RNA secondary structure modeling is crucial for understanding RNA function.
• Chemical probing, like SHAPE, coupled with thermodynamic data, enables RNA structure prediction.
• Modeling large and complex eukaryotic RNAs, such as ribosomal RNAs (rRNAs), presents a significant challenge.

Purpose of the Study:

• To validate and improve experimental approaches for modeling large RNA transcripts.
• To assess the structural integrity of protein-free human 18S and 28S ribosomal RNAs (rRNAs).
• To compare the structural determinacy of eukaryotic rRNAs with prokaryotic counterparts.

Main Methods:

• High-quality SHAPE (Selective 2 '-Hydroxyl Acylation Measured by Proportionality) data collection for human 18S and 28S rRNAs.
• Computational modeling of rRNA secondary structures guided by SHAPE data.
• Analysis of structural determinacy using low-SHAPE and low-entropy (lowSS) metrics.

Main Results:

• SHAPE-directed models for human rRNAs showed poor agreement with accepted structures.
• Only ~40% of human rRNA sequences formed well-determined lowSS secondary structures, compared to ~75% in *Escherichia coli*.
• Regions of human rRNA that did form well-determined lowSS structures were accurately modeled using SHAPE data.

Conclusions:

• Eukaryotic rRNAs are significantly more unfolded than prokaryotic rRNAs, likely due to higher protein dependence in eukaryotes.
• Despite overall unfolding, well-defined substructures within eukaryotic rRNAs can be identified and accurately modeled.
• SHAPE-directed folding remains a powerful tool for modeling the structured regions of complex RNAs.