TnpBs: Selfish Genes Unlock New Opportunities in Genome Editing
Transposons are “selfish” genetic elements that have evolved by excising themselves from DNA and inserting themselves elsewhere in the genome. Some transposon families include a programmable endonuclease, TnpB, thought to be an ancestor of Cas enzymes used in CRISPR-Cas genome editing. These ancient origins may hold the key to unlocking new possibilities in genetic modification, both in animal and plant systems.
Some TnpBs are less than half the size of CRISPR-Cas enzymes (~400 amino acids compared to >1000), meaning they aren’t restricted by many of the same physical barriers. In plants, virus-induced genome editing (VIGE) offers many opportunities in the genetic modification of plants, with heritable germline edits skipping the need for time- and resource-intensive tissue culture systems. The problem with viral delivery is that viruses often have a finite carrying capacity, and cannot reliably transport large systems like CRISPR-Cas, a problem circumvented with TnpBs. In addition to their potential in plants, viral delivery holds promise for animal systems through adeno-associated viruses (AAVs), which can deliver genome-editing machinery to cells without being able to replicate itself. These exciting applications are fueling further research in TnpBs, which requires a developed understanding of their mechanisms.
Despite their evolutionary origins, TnpBs operate very differently from Cas enzymes. In order for the transposon to move across the genome, it must “jump” to a new site, a process that involves excision from the original site. This excision is performed by TnpA, a transposase that recognizes the beginning and end of the transposon and cuts at each sequence. The excised transposon can then insert itself into a new locus in the genome, but the original copy of the transposon must be restored.
To restore the transposon in the edited strand of DNA, TnpB initiates homologous recombination, a cellular mechanism that copies genetic information from one strand to another. This requires TnpB to cut the DNA at the excision site, which is made possible with guidance from the sequence encoded within the TnpB’s “right end element RNA” (reRNA), enabling TnpB to target a transposon-associated motif (TAM), a programmable site with a unique sequence requirement.
By locating the TAM and target site, the reRNA can position TnpB to make a double-stranded DNA (dsDNA) break at its target site, where the TnpB was originally excised. Recognizing that DNA damage has occurred, the cell performs homologous recombination, using DNA from the complementary strand as a reference to repair any potential mistakes. With a transposon in both strands of DNA, the transposon can excise itself once again, and the cycle continues.
The ability of TnpBs and their reRNA guides to recognize specific motifs and make precise, programmable cuts is what powers their potential as a novel tool in genome editing. However, with the RNA-guided endonuclease activity of TnpB only being experimentally characterized in 2021, this field remains new. As such, TnpBs had not been naturally evolved for high editing efficiency enough to compete with other systems like CRISPR-Cas. For implementation in modern biotechnology, TnpB systems must be improved, an initiative led by Brittney Thornton and Rachel Weissman of the Savage and Doudna labs at UC Berkeley’s Innovative Genomics Institute. Through deep mutational scanning (DMS), these researchers evaluated the effects of changing every individual amino acid of TnpB.
Favorable mutations were identified by using deep mutational scanning libraries of TnpB, encompassing nearly all single–amino acid substitutions in the protein and single–nucleotide mutations in the reRNA. These pooled variant libraries were transformed into a yeast reporter strain, and they measured performance of each variant by programming TnpB to repair an essential gene, based on an assay previously used to improve CRISPR-Cas9. Edited yeast cells can survive and grow into colonies, while unedited cells do not proliferate. By sequencing DNA from the colonies that grew, they counted how many times each variant appeared in edited yeast cells. By scoring TnpB mutants against the unmodified (WT) TnpB, mutations can be associated with a positive or negative effect on activity.
This is unlike more typical approaches to protein engineering, such as error-prone PCR (epPCR), which creates random mutations. While epPCR allows for researchers to select for favorable mutations, which is crucial for engineering better TnpBs, DMS provides a more comprehensive dataset. DMS revealed the effects, both positive and negative, of changing every individual amino acid, allowing researchers to identify patterns in performance. This is especially valuable as research on this system remains early, providing information on how the protein works. This decision, in addition to choosing well-characterized TnpB variant ISDra2, provided researchers with the information required to draw confident conclusions. “We were about to do this big screening assay and we wanted to be very rigorous about being able to interpret mutational effects,” explained Dr. Brittney Thornton, co-author of the study. Having a good understanding of the biochemical structure enabled them to identify positive and negative controls, allowing them to ensure that the enriched mutations were rational. Along with increasing confidence in their results, this strong foundation enabled them to make hypotheses about other gain- or loss-of-function mutations.
The team evaluated the effects of 93% of all possible substitutions in the ISDra2 TnpB protein (7,611 of 8,140 amino acids), finding that positively-charged amino acids were typically enriched in regions that physically interact with DNA. This discovery is consistent with that of previous studies on CRISPR-Cas12 enzymes, as negatively charged protein components are repelled from DNA, which is also negatively charged. Remarkably, 20% of single amino acid mutations resulted in an increase in activity compared to WT TnpB. With a one-in-five chance of a random amino acid mutation enhancing TnpB activity, TnpB’s performance is theorized to be under negative selection. Considering TnpB’s role as a transposon-associated endonuclease, it makes sense that evolution might be holding it back: excessive double-stranded breaks is not advantageous if it compromises host fitness. A hyperactive transposon risks disrupting essential genes, which would harm its host, reducing its evolutionary success.
In addition to the TnpB protein, a DMS was also conducted on the reRNA that directs the endonuclease to its target site. This comprehensive assay allowed researchers to visualize patterns relating to its higher-order nucleic acid structure in performance, including two small regions of the reRNA that showed stark enrichment. Based on existing studies of the reRNA’s structure, these regions have been proposed to act as a “hinge,” regulating TnpB activity. It is hypothesized that the reRNA can regulate TnpB nuclease activity, a state determined by the hinge. Therefore, mutations in this region may leave TnpB in a state free to make more edits.
From their library of mutations, researchers chose 33 of the best mutations and combined them to create about 5,000 TnpB mutants. After more assays in yeast, five of the best mutants were selected for further testing in plants, using model organism Nicotiana benthamiana. All variants exhibited editing efficiencies 4-40 times that of WT TnpB, but two top candidates were chosen for further tests, referred to as TnpB-KYLI and TnpB-VGIRL. Interestingly, complementing these enhanced TnpB mutants with reRNAs with hinge mutations did not improve editing efficiency. Although the engineered reRNAs would theoretically enable higher editing rates, it is thought that the combination of this structural change, in addition to the 4-5 mutations made in the enhanced TnpB variants, compromised stability of the TnpB-reRNA complex.
Paired with WT reRNA, TnpB-KYLI and TnpB-VGIRL outperformed other novel engineered small RNA-guided endonucleases at editing in N. benthamiana, and were subsequently used in a more recent study in collaboration with the Dinesh-Kumar lab at UC Davis. Using these enhanced mutants for VIGE in N. benthamiana, researchers found heritable germline editing in seeds infected with TnpB-KYLI, but not WT TnpB or TnpB-VGIRL. This advancement demonstrates the potential of TnpB research, both in its current capabilities and scientists’ ability to spur it forward. By adapting nature’s tools to humanity’s needs, this exciting field promises to advance human knowledge while inventing new possibilities.
I was fortunate to meet Dr. Thornton at a poster presentation at the Innovative Genomics Institute in October. I was excited to see her research on TnpBs, which I had just covered for a previous article about their potential for plant genome editing. I’ve been able to speak with her a few times about her research journey and work on this project, which have been really engaging and informative conversations. It was really great to be able to talk directly to those involved in the project, because I had a lot of questions that weren’t addressed in the literature I read. Visualizing the genetic structure and activity of TnpBs made it a lot easier to understand what mechanisms were at work and how they relate to the development of TnpBs as a genome-editing tool. Hopefully I was able to translate that with the diagrams embedded across this article.
From our discussions, it looks like this field is really rapidly expanding, which is exciting both because of the new directions being explored and the potential recognized by the researchers involved. I can’t wait to see what is coming next for this field, so stay tuned to learn with me!
Check out the main paper featured in this article on mutational scanning of TnpBs: Mutational scanning of TnpB reveals latent activity for genome editing
Also see how TnpB-KYLI and VGIRL (also called eTnpBc and eTnpBe respectively) were used for plant genome editing: High-efficiency, transgene-free plant genome editing by viral delivery of an engineered TnpB