Biology subjects

Spellmon, N.

Publications and source records attributed to Spellmon, N..

3 recordsLinked to original sources

bioRxiv2026

Cryo-EM of a nucleotide-polymerizing ribozyme enables its predictive improvement

Ribozymes capable of self-replication from nucleotides would have been central to the hypothesized RNA World. The leading laboratory models for such molecules were converted from a class I ligase by in vitro evolution but then developed without 3D structures. Here, scaffolded cryo-EM of the substrate-free tC19Z RNA polymerase ribozyme at 3.1 [A] resolution shows how this conversion was achieved. An accessory domain evolved from random sequence grips the ancestral ligase through a loop-loop contact, a seam of magnesium ions, and a six-base stack, and rebuilds the ligases substrate binding site from different residues of its own. A previously unrecognized pairing, present before substrate binds, sequesters the 5' end that must otherwise pair with the template. Compensatory mutations to the ribozyme and template, designed to break this ectopic pairing, increase the extension rate. These results suggest that accelerating RNA structure determination may speed progress toward nucleotide-based self-replication.

biochemistry↗

bioRxiv2026

De novo design of RNA pseudoknots with deep learning

RNA design has been hindered by the limited accuracy of 3D structure prediction. Here, we show that intricate RNA structures can be generated with current deep learning tools through accurate de novo design of pseudoknot secondary structures. In an Eterna competition involving 57 pseudoknots, generative AI methods matched experienced human designers in solving most blind challenges, evaluated by single-nucleotide-resolution chemical mapping, compensatory mutagenesis, and cryogenic electron microscopy. AI-generated molecules with accurate secondary structures formed well-ordered 3D folds stabilized by noncanonical tertiary interactions not modeled during design. Success was guided by an RNet foundation model trained on prior chemical mapping data, suggesting that some difficult RNA design tasks may be tractable without first solving RNA 3D structure prediction.

biophysics↗

bioRxiv2024

Structure of the SMYD2-PARP1 Complex Reveals Both Productive and Allosteric Modes of Peptide Binding

Allosteric regulation allows proteins to dynamically respond to environmental cues by modulating activity at sites away from the catalytic center. Despite its importance, the SET-domain protein lysine methyltransferase superfamily has been understudied. Here, we present four crystal structures of SMYD2, a unique family member with a MYND domain. Our findings reveal a novel allosteric binding site with high conformational plasticity and promiscuity, capable of binding peptides, proteins, PEG, and small molecules. This site exhibits positive cooperativity with substrate binding, influencing catalytic activity. Mutations here significantly alter substrate affinity, changing the enzymes kinetic profile. Specificity studies show interaction with PARP1 but not histones, suggesting targeted regulation. Interestingly, this sites function remains unaffected by active site changes, indicating unidirectional mechanisms. Our discovery provides novel insights into SMYD2s biochemical regulation and lays the foundation for broader research on allosteric control in lysine methyltransferases. Given SMYD2s role in various cancers, this work opens exciting avenues for designing specific allosteric inhibitors with reduced off-target effects.

biochemistry↗