loci encoding lncRNA aren't genes? How would you define a gene? Notoriously tricky problem.

Employing "a much larger ribozyme" is always the solution to any problem, in my opinion.

Thank you so much for sharing! I'm glad you like it!

Thank you! Yes, absolutely! We wouldn't know what to do without it. I strongly recommend people check it out! www.rnacanvas.app

Yes, that's correct, and honestly I was surprised with how well it worked, too. I believe the trick was paying close attention to sequence/structure motif conservation which seems to hold sufficient information to reconstruct the whole ribozyme class.

I'm grateful to my student Lukas, who did a lot of the early characterization of Arq.I2, and of course @mutschlerlab.bsky.social for his mentorship and making this project possible! And ofc @crc392molevo.bsky.social!

I will be answering any questions, so feel free to ask if you have any! (11/11)

This project was my brainchild and I'm overjoyed to finally see it published!

I want to thank my co-1st author: @noemi-n.bsky.social for all her hard work in seeing the project through until the end! I expect a lot of cool upcoming work from her, so follow her if you like group II introns! (10/11)

This work serves as a proof of concept for the viability of complex ncRNA design through inverse folding, and we hope it paves the way for the computational design of bespoke G2Is and possibly other complex ncRNAs with traits that are favorable for biotech or therapeutic applications. (9/11)

Surprisingly, we saw modest GFP expression compared to a control with an inactive intron. G2Is tend to silence host genes by cleaving the mRNA, so lowered expression is not unexpected. These results indicate that the protein-free intron is active in E. coli cells. (8/n)

According to conventional wisdom, G2Is require protein cofactors to aid in self-splicing under intracellular conditions, hindering their application as protein-free genome editors.

We designed an assay for splicing-dependent fluorescent protein expression — a GFP gene interrupted by Arq.I2 (7/n)

Unlike typical de novo designed ribozymes, Arq.I2 outperforms most wild type variants characterized thus far! It reacts with fast kinetics, having a rate of splicing comparable to those of the fastest known natural G2Is! (6/n)

All three candidates were active in vitro, some even outperforming a variant of a natural intron.
Arq.I2 is a very unusual intron — it requires extreme measures to denature it for PAGE, and remains catalytically active in the absence of monovalent cations. (5/n)

In our latest paper we combined the inverse folding algorithm aRNAque and rational design to generate unique synthetic G2Is.
aRNAque's evolutionary algorithm was able to finetune intron folding, resulting in unusually stable structures. (4/n)

Protein-free G2Is were recently found to hydrolyze dsDNA, showing potential as RNA-only genome editors (see my previous thread). However, their complexity poses a challenge to their de novo design, as only short ribozymes have ever been generated using inverse folding algorithms. (3/n)

G2Is are self-splicing retroelements found in all domains of life, and are the ancestors of the spliceosome.
Owing to dozens of tertiary interactions and a large conformational change between the steps of catalysis, they are among the largest and most complex ribozymes. (2/n)