Unquestionably, we will emerge from this revolutionary period with modified views of components of cells and how they operate, but only, however, to await the emergence of the next revolutionary phase that again will bring startling changes in concepts. —Barbara McClintock, Nobel laureate (1983)

 “Everything in the last 14 years of genome editing has been based on CRISPR. We have been whipping this horse for a decade and a half, but we need more programmable functions with complexity beyond the molecular scissors that cut RNA and DNA.” So says Patrick Hsu, PhD, co-founder of the Arc Institute, in whose lab the next revolutionary phase of genome engineering may have just been unearthed, even as the CRISPR revolution has barely begun.  

In January of this year, researchers in Hsu’s laboratory (he is also assistant professor of bioengineering and Deb Faculty Fellow at the University of California, Berkeley) posted a preprint on bioRxiv in which they claim to have discovered a new class of natural single-effector RNA-guided systems.1 That story, now peer reviewed, is published online today in Nature.  

Fig 1. Programmable RNA-guided tools. Three generations of programmable RNA-guided tools—RNAi (left, blue); CRISPR-Cas (middle, purple); and bridge RNAs (right, green)—critical for the new frontier of genome design. [Credit: Durrant, Perry et al.]

These systems retain the key property of programmability from RNAi and CRISPR, while enabling large-scale genome design beyond RNA and DNA cleavage. These modular, bi-specific bridge RNAs can be reprogrammed to enable sequence-specific fundamental DNA rearrangements, potentially accelerating the advancement of genome design (Fig 1).2 

“This is a much more complex molecular machine,” Hsu continues. “We’re excited about the potential of this for eventually achieving chromosome-scale genome engineering, where you can do long-range insertions, deletions, and genome translocations.” 

Complex genome assembly 

Hsu is a self-described technologist—he creates the genome engineering and biological design tools of tomorrow. As a graduate student at Harvard University, he worked with his mentor Feng Zhang, PhD, building some of the foundational components of CRISPR as a genome engineering tool.3 But Hsu came to realize that manipulating RNA might be a more flexible technique than making permanent, and sometimes unintended, changes to the genetic code. In 2018, Hsu and his team, which included Arc Institute co-founder Silvana Konermann, PhD, developed CasRx as a programmable RNA-binding module for efficient targeting of cellular RNA, enabling a general platform for transcriptome engineering and future therapeutic development with RNA targeting CRISPR.4  

Hsu believes that one of the greatest challenges today is the manipulation of eukaryotic genomes, particularly the integration and manipulation of large, multi-kilobase (kb) DNA sequences, which limits the rapidly growing fields of synthetic biology, cell engineering, and gene therapy. According to Hsu, who is quite the science historian, the prevailing modern genome editing method has been stuck on a road paved with milestones that began with the initial work of Mario Capecchi, PhD, in 1980 on the insertion of DNA into mammalian cells,5 Martin Evans PhD’s work in 1986 on chimeric mice,6 and then work from both Oliver Smithies, PhD, and Capecchi to successfully create specific modifications in the genomes of mice.7,8   

“That was generally an attempt to do an insertion reaction, which is only one way to do genome editing,” Hsu told me. “Funnily enough, what happens today with base editing or prime editing—while exciting—is arguably even smaller in scope. You’re making single-nucleotide polymorphism changes or tens of bases, rather than the multigene-sized cassettes that we originally [envisioned]… Generally, all of this has been small-scale single-locus changes.” 

Hsu argues that CRISPR-Cas molecular scissors require a complex multistep process to make an edit that, in chemistry terms at least, has very low purity. “You mostly get indels, and only a small amount of time you get the edit, which then leads to chromosomal translocations and large lesions,” said Hsu.  

“You’re trying to create this chemical reaction that’s just not very predictable. Certainly, there are highly optimized cases like ex vivo T-cell engineering, where billions of dollars have been invested into this problem to get it to work right. [Scientists] will hang their hat on that and say that CRISPR is high efficiency and specificity. But generalizing this success to other cell types or in vivo has been really hard—it’s a fundamental mechanistic limitation.” 

Hsu is not alone in this view. In 2019, Harvard Medical School geneticist George Church, PhD, memorably referred to CRISPR as a “blunt axe” that performs “genome vandalism.” Instead of looking to take a major leap forward, Hsu said the genome engineering field is in a battle over bragging rights that centers around the size, toxicity, specificity, and efficiency of their nucleases.  

Some recently developed gene editing assemblies, such as PASTE and PASSIGE (both developed by former colleagues of Hsu’s at the Broad Institute), are showing considerable promise in their own right, although they require several distinct components to be delivered to a given cell.9   

By contrast, Hsu is trying to reframe the conversation completely. Hsu, who was already working on a solution to evolve the genome engineering field, did what many talented scientists have done in the past—he turned to nature for inspiration. For example, during conception, there’s a large amount of recombination and genomic rearrangements from crossover events at chromosomal scale between genes from the mother and father that results in the unique individuals that we are.  

Another example is Deinococcus radiodurans, a bacterium that is extremely resistant to radiation and other environmental stresses. When the organism is being battered by radioactivity, their genomes can shatter, but they always find a way to reassemble a complete genome. And then there are mobile genetic elements (MGEs), which brings us to the great Barbara McClintock, PhD. 

About 80 years after Gregor Mendel worked with peas in an abbey garden in (what is now) the Czech Republic to describe the transmission of genetic traits (before anyone knew genes existed), McClintock began experiments on maize kernels at the Cold Spring Harbor Laboratory that would lead to her profound discoveries on transposable elements.10,11 McClintock observed that genetic elements can change position on a chromosome, causing nearby genes to become active or inactive, and that this correlated with the redistribution of genetic traits in maize as well as other organisms.12 

In her 1983 Nobel Prize lecture, McClintock spoke about genetic shocks created by MGEs that create selective pressures and, thus, new gene functions—even the origin of new species.13 “[McClintock] was one of the original people who discovered these elements, one of the first people that realized that they’re actually really powerful driving forces for genetic diversification in new functions,” said Hsu. “We’ve just been fascinated by these transposons and genomic elements and what else could be out there.” 

That framework got Hsu to begin examining large serine recombinases (LSRs). “The reason we did our LSR research in the first place, was not really to find recombinases per se—it was to solve this core technological problem of modern gene editing.” 

The bridge RNA recombinase mechanism 

A few years ago, two members of the Hsu lab, graduate student Nicholas Perry and computational biologist Matt Durrant, PhD, were working on LSRs (Fig 2). One of the perks of working in the non-profit Arc Institute is that it affords scientists no-strings-attached, multi-year funding.14 Perry and Durrant decided to sift through a huge genomic and metagenomic database that Durrant had been compiling for MGEs (Box 1).  

Fig 2 .Arc Institute study co-authors Patrick Hsu, Nicholas Perry, and Matt Durant. [Ray Rudolph.]

Upon investigating a family of cut-and-paste MGEs called IS110s, which all had a gene encoding a RuvC-like domain—one of the key nuclease domains in the Cas9 nuclease—the Arc team wondered if it might encode a whole RNA-guided transposase system.15,16 This made sense to Durrant, as IS110 elements scarlessly excise themselves from the genome and generate a circular form as part of their transposition mechanism (Fig. 3)

Fig 3. IS110 family elements are cut-and-paste mobile genetic elements. IS110 elements utilize a recombinase to scarlessly excise out of their genomic context, yielding a dsDNA circular form that is inserted into specific genomic target sequences (blue) such as repetitive extragenic palindromic elements. Recombination of the circular form and the target is centered around a short core sequence (green diamonds), which appears as a direct repeat immediately flanking the inserted element. The intervening sequences between the cores and the recombinase coding sequence (gray) are defined as the left and right non-coding ends (orange). [Durrant, Perry et al.]

“Patrick threw around the idea that maybe there’s some kind of RNA-guided transposase out there,” Durrant recalled. “At that point, I had just been staring at sequences for so long that I felt like I had built up an intuition for what would be worth pursuing.” 

Hsu’s and Durrant’s hypothesis sent Perry on a quest to discover whether there could be some sort of RNA that existed, a pre-requisite for any sort of RNA-guided transposase system. Their first clue came from secondary structure analysis of a particular IS110 sequence, called IS621, which was predicted to contain an RNA with a 5′ stem loop and two large internal loops. Perry set out to demonstrate that an RNA was expressed from this sequence. 

Meanwhile, Durrant circled back to an idea that Hsu had pushed him to implement called covariation analysis to predict base-pairing interactions between the non-coding RNA (ncRNA) and the target or donor DNA.17 This comparative analysis method asks whether pairs of nucleotides change in tandem at specific positions of aligned DNA and RNA, which would indicate evolutionary pressure to conserved base-pairing interactions between ncRNA positions and target or donor positions. Projecting this covariation pattern onto the canonical IS621 sequence and ncRNA secondary structure, Durrant saw that the first internal loop may base-pair with the target DNA.  

“It was a really messy initial analysis… but we got lucky enough that it worked,” said Durrant. “We saw this signal of bits of the non-coding ends that co-vary with the target. It was the most exciting moment of my whole career, when it became clear that these are probably programmable.” 

Durrant immediately messaged Perry, who was at a CRISPR conference in Boston, saying they had to talk immediately. Perry hurriedly left the meeting and covertly called Durrant, wary of any potential eavesdroppers who might pick up on their findings about RNA-guided transposases. Although the team gave some thought to publishing the computational results alone, Durrant said there was “some sliding scale of confidence and doubt… We tried to deeply understand how everything worked.” 

For the next few months, Perry was buried in experiments to try to assemble pieces of the puzzle that had been predicted by Durrant’s computational analysis. On a Sunday morning, which happened to be his birthday, Perry was looking through data when he got confirmation for an experiment showing that a promoter on the circular form of IS621 expressed an encoded ncRNA that formed a functional complex with a transposase. 

“I spent probably six months before getting any conclusive positive data about the expression of a non-coding RNA, observing them transpose and showing that they worked in an orthogonal system, trying to engineer the system,” said Perry. “If we couldn’t even get these to function in the cell type that they’re native in, then how could we ever learn anything more about them? You can wish that they’re programmable, but if there’s no observable function, does it really matter?” 

Perry was able to confirm a mechanism for a programmable target loop, and a short time later, they repeated the whole process for identifying and confirming that the second loop was a programmable donor loop. This suggested to Perry and Durrant that the RNA acts as a “bridge” between target and donor DNA to enable recombination by the IS621 recombinase (Box 2)

Then came the eureka moment, where the entire picture of their computational and experimental data came together. Perry and Durrant had discovered a single-effector recombinase system that uses a bridge RNA with two distinct binding loops that can be independently reprogrammed to bind and recombine diverse DNA sequences (Fig. 4). “We realized it was programmable on both [the donor and target] ends, which was completely unprecedented,” said Perry. 

Fig 4. A bispecific bridge RNA recognizes target and donor DNAs The IS621 bridge RNA contains two internal loops: the target-binding loop (blue) and donor-binding loop (orange). The target-binding loop comprises two key regions that base-pair with the top and bottom strands of the target DNA, respectively: the left target guide (LTG) base-pairs with the left side of the bottom strand of the target DNA (left target; LT), while the right target guide (RTG) base-pairs with the right top strand of the target DNA (right target; RT). The donor-binding loop has an analogous architecture, with a left donor guide (LDG) base-pairing with the bottom strand of the left donor (LD) and a right donor guide (RDG) base-pairing with the top strand of the right donor (RD). Importantly, the core dinucleotide is included in every one of the base-pairing interactions (LTG-LT, RTG-RT, LDG-LD, and RDG-RD), resulting in an overlap between the right top and left bottom strand pairings and suggesting a key role for bridge RNA-core interactions for recombination. [Durrant, Perry et al.]

The team went on to demonstrate that the modularity of each loop of the bridge RNA can facilitate the recombinase system to execute sequence-specific insertions, inversions, and excisions. This meant that the bridge RNA-guided single effector system provided “a unified, programmable, and modular mechanism for the fundamental DNA rearrangements required for genome design,” says Hsu. “We discovered a conceptually distinct mechanism of RNA-guided self recognition for a mobile genetic element and capitalized upon this mechanistic feature to enable a new method of genetic engineering.” 

A GUI for genome engineering 

Since the early days of bacterial artificial chromosomes (BACs) and yeast artificial chromosomes (YACs), assembling large sections of DNA has been technically incredibly challenging. As the field has moved towards making synthetic minimal genomes, the field has been limited to relatively short DNA synthesis and assembly techniques. With the concept of genome design in mind, Hsu explains the significance of these bridge RNAs in computer terms—they’re basically like a program that allows the user to install and uninstall packages. 

“The Xerox Alto was the first computer ever sold with a mouse and a graphical user interface (GUI),” said Hsu. “It was invented just off the street from where Arc [Institute] is today at Xerox PARC. It gave humans for the first time a simple and intuitive way to interact with information. Guide RNAs act like that mouse cursor to interact with nucleic acids in a large genome. What we’ve been doing so far is basically punching individual nucleotides and changing them, like punch-card programming. We want something that can operate at a much higher level of abstraction to design genomes. That’s where all of this is going.” 

Nature has invented it, and there are probably many more programmable systems. And Hsu will keep on looking to see what other biotechnology doors he can unlock. 


  1. Durrant MG, Perry NT, Pai JJ, et al. Bridge RNAs Direct Modular and Programmable Recombination of Target and Donor DNA. 2024;2024.01.24.577089; doi: 10.1101/2024.01.24.577089.
  2. Herrera RJ, Garcia-Bertrand R, Salzano FM. Genomes, Evolution, and Culture: Past, Present, and Future of Humankind. Wiley; 2016.
  3. Hsu PD, Scott DA, Weinstein JA, et al. DNA targeting specificity of RNA-guided Cas9 nucleases. Nat Biotech 2013;31:827-32. doi: 10.1038/nbt.2647
  4. Konermann S, Lotfy P, Brideau NJ, et al. Transcriptome Engineering with RNA-Targeting Type VI-D CRISPR Effectors. Cell 2018;173(3):665-676.e14; doi: 10.1016/j.cell.2018.02.033.
  5. Capecchi MR. High efficiency transformation by direct microinjection of DNA into cultured mammalian cells. Cell 1980;22(2, Part 2):479–488; doi: 10.1016/0092-8674(80)90358-X.
  6. Robertson E, Bradley A, Kuehn M, et al. Germ-line transmission of genes introduced into cultured pluripotential cells by retroviral vector. Nature 1986;323(6087):445–448; doi: 10.1038/323445a0.
  7. Doetschman T, Gregg RG, Maeda N, et al. Targetted correction of a mutant HPRT gene in mouse embryonic stem cells. Nature 1987;330(6148):576–578; doi: 10.1038/330576a0.
  8. Thomas KR, Capecchi MR. Site-directed mutagenesis by gene targeting in mouse embryo-derived stem cells. Cell 1987;51(3):503–512; doi: 10.1016/0092-8674(87)90646-5.
  9. Yarnall MTN, Ioannidi EI, Schmitt-Ulms C, et al. Drag-and-drop genome insertion of large sequences without double-strand DNA cleavage using CRISPR-directed integrases. Nat Biotechnol 2023;41(4):500–512; doi: 10.1038/s41587-022-01527-4.

9A. Pandey S, Gao XD, Krasnow NA, et al. Nat Biomed Eng. Efficient site-specific integration of large genes in mammalian cells via continuously evolved recombinases and prime editing. June 10, 2024. doi: 10.1038/s41551-024-01227-1 

  1. McClintock B. The origin and behavior of mutable loci in maize. Proceedings of the National Academy of Sciences 1950;36(6):344–355; doi: 10.1073/pnas.36.6.344.
  2. McClintock B. INDUCTION OF INSTABILITY AT SELECTED LOCI IN MAIZE. Genetics 1953;38(6):579–599; doi: 10.1093/genetics/38.6.579.
  3. McClintock B. Some Parallels Between Gene Control Systems in Maize and in Bacteria. The American Naturalist 1961;95(884):265–277; doi: 10.1086/282188.
  4. McClintock B. The Significance of Responses of the Genome to Challenge. Science 1984;226(4676):792–801.
  5. Grinstein JD. Ahead of the Curve: The Arc Institute Aims to Reshape the Biomedical Research Landscape | GEN Biotechnology. n.d. Available from: https://www.liebertpub.com/doi/10.1089/genbio.2022.29017.jgr [Last accessed: 4/18/2024].
  6. Higgins BP, Carpenter CD, Karls AC. Chromosomal context directs high-frequency precise excision of IS492 in Pseudoalteromonas atlantica. Proceedings of the National Academy of Sciences 2007;104(6):1901–1906; doi: 10.1073/pnas.0608633104.
  7. Partridge SR, Hall RM. The IS1111 Family Members IS4321 and IS5075 Have Subterminal Inverted Repeats and Target the Terminal Inverted Repeats of Tn21 Family Transposons. Journal of Bacteriology 2003;185(21):6371–6384; doi: 10.1128/jb.185.21.6371-6384.2003.
  8. Ekeberg M, Lövkvist C, Lan Y, et al. Improved contact prediction in proteins: Using pseudolikelihoods to infer Potts models. Phys Rev E 2013;87(1):012707; doi: 10.1103/PhysRevE.87.012707.
  9. Durrant MG, Fanton A, Tycko J, et al. Systematic discovery of recombinases for efficient integration of large DNA sequences into the human genome. Nat Biotechnol 2023;41(4):488–499; doi: 10.1038/s41587-022-01494-w.
  10. Fire A, Xu S, Montgomery MK, et al. Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 1998;391(6669):806–811; doi: 10.1038/35888.
  11. Jinek M, Chylinski K, Fonfara I, et al. A Programmable Dual-RNA–Guided DNA Endonuclease in Adaptive Bacterial Immunity. Science 2012;337(6096):816–821; doi: 10.1126/science.1225829.
  12. Cong L, Ran FA, Cox D, et al. Multiplex Genome Engineering Using CRISPR/Cas Systems. Science 2013;339(6121):819–823; doi: 10.1126/science.1231143.
  13. Komor AC, Kim YB, Packer MS, et al. Programmable editing of a target base in genomic DNA without double-stranded DNA cleavage. Nature 2016;533(7603):420–424; doi: 10.1038/nature17946.
  14. Anzalone AV, Randolph PB, Davis JR, et al. Search-and-replace genome editing without double-strand breaks or donor DNA. Nature 2019;576(7785):149–157; doi: 10.1038/s41586-019-1711-4.
  15. Anzalone AV, Koblan LW, Liu DR. Genome editing with CRISPR–Cas nucleases, base editors, transposases and prime editors. Nat Biotechnol 2020;38(7):824–844; doi: 10.1038/s41587-020-0561-9.
  16. Tou CJ, Kleinstiver BP. Recent Advances in Double-Strand Break-Free Kilobase-Scale Genome Editing Technologies. Biochemistry 2023;62(24):3493–3499; doi: 10.1021/acs.biochem.2c00311.
  17. Prostova M, Kanevskaya A, Panteleev V, et al. DNA-targeting short Argonautes complex with effector proteins for collateral nuclease activity and bacterial population immunity. Nat Microbiol 2024;1–14; doi: 10.1038/s41564-024-01654-5.


Previous articleGene Variant Linked to Ancestry-Related Disparity in Diabetes Complications
Next articleGeoVax Advances COVID-19 Vaccine with Up to $45M from BARDA