Directed evolution of compact RNA-guided nucleases for enhanced activity in mammalian cells

This study employs directed evolution in human cells to develop compact, highly active RNA-guided nucleases (Cas12f1Super and TnpBSuper) that overcome the size limitations of AAV delivery while significantly enhancing editing efficiency for potential therapeutic applications.

The Gist

Imagine your DNA is a massive, intricate instruction manual for building and running a human body. Sometimes, there are typos in this manual that cause diseases. Scientists have developed a tool called CRISPR that acts like a pair of "molecular scissors" to find those typos and cut them out so the body can fix them.

However, there's a big problem with the best scissors we have (like the famous Cas9). They are too big.

Think of trying to deliver a giant refrigerator (the big scissors) into a tiny studio apartment (a human cell) using a small delivery truck (a virus vector called AAV). The truck simply can't fit the fridge inside. Because of this, scientists have been looking for "mini-fridges"—smaller, more compact scissors that can fit in the truck.

They found some candidates: tiny scissors called Cas12f and TnpB. But here's the catch: while they fit in the truck, they are clumsy. They are slow, they miss their target often, and they don't cut very well inside human cells.

The Solution: "Evolution in a Test Tube"

The researchers in this paper decided to fix these clumsy scissors using a process called Directed Evolution.

Think of this like a high-stakes talent show or a survival game for proteins:

1. The Setup: They took the genes for these tiny scissors and made millions of slightly different versions of them. It's like taking a recipe for a cake and making millions of cakes where you changed one ingredient in each one (maybe too much sugar, maybe less flour, maybe a different spice).
2. The Challenge: They put these millions of "mutant" scissors into human cells. The cells had a broken lightbulb (a gene that makes them glow green) that could only be fixed if the scissors cut the DNA in the exact right spot and the cell's repair crew fixed it perfectly.
3. The Selection: Only the cells where the scissors did a great job would light up green. The scientists used a machine to sort through millions of cells and grab only the glowing ones.
4. The Breeding: They took the DNA from those glowing cells, mixed the best mutations together, and made a new, even better generation of scissors. They repeated this process over and over.

The Result: "Super Scissors"

After several rounds of this "survival of the fittest," they created two new champions:

• Cas12f1Super
• TnpBSuper

These new versions are 11 times more efficient than the original clumsy ones. They are like upgrading from a dull, rusty pocketknife to a laser-guided scalpel.

Why is this a big deal?

• They fit the truck: They are still small enough to be delivered by the AAV virus, which is crucial for treating diseases in the human body.
• They are precise: They cut the DNA exactly where they are supposed to, without accidentally cutting other parts of the manual (which could cause cancer or other problems).
• They work everywhere: The scientists tested them in human cells, in immune cells (T-cells), and even in live mice. In the mice, they successfully edited the liver to lower cholesterol levels, proving they work in a living animal.
• They are versatile: They can even be turned into "base editors" (which change a single letter in the DNA code without cutting the whole strand), and they work 10 times better than previous versions.

The Analogy Summary

Imagine you are trying to fix a specific typo in a 1,000-page book, but you can only use a tiny, handheld tool that fits in a pocket.

• Old Tools: The big scissors (Cas9) are too heavy to fit in your pocket.
• New Tools (Original): The tiny scissors (Cas12f/TnpB) fit in your pocket, but they are made of cheap plastic and break easily.
• This Paper's Tools: The scientists took those cheap plastic scissors, ran them through a "training camp" inside human cells, and evolved them into titanium, laser-guided scissors that fit in your pocket, never miss a page, and can fix the book in record time.

This breakthrough opens the door to treating many more genetic diseases because we finally have tools that are both small enough to deliver and powerful enough to work.

Technical Summary

1. Problem Statement

CRISPR-Cas technology holds immense promise for treating genetic diseases, particularly through precise sequence replacement (Homology-Directed Repair, HDR). However, the most effective nucleases, such as S. pyogenes Cas9 (SpCas9), are too large to be packaged into Adeno-Associated Virus (AAV) vectors, the gold standard for in vivo gene therapy delivery.

While smaller alternatives like Cas12f (e.g., Un1Cas12f1) and TnpB (e.g., ISDra2 TnpB) fit within AAV packaging limits, they currently suffer from significantly lower editing efficiencies in mammalian cells compared to larger Cas enzymes. Previous attempts to improve these enzymes via rational design or directed evolution in non-mammalian systems (bacteria, phage) have often failed to translate to human cells due to differences in cellular biology (e.g., chromatin structure, DNA repair pathways, and protein folding). Furthermore, deep mutational scanning (DMS) in human cells is typically limited to single-point mutations, missing the potential synergistic effects of multi-mutation combinations.

2. Methodology

The authors developed a mammalian cell-based directed evolution platform to select for variants with enhanced HDR activity directly in the relevant biological context.

• Reporter System: They engineered a fluorescent HDR reporter in HEK293T cells. The reporter contained a disrupted EGFP gene (lacking a start codon) preceded by a protospacer and PAM/TAM sequence. Successful HDR using a donor ssODN (containing the start codon) restored EGFP expression.
• Directed Evolution Workflow:
  1. Library Generation: Error-prone PCR (epPCR) was used to generate libraries of ~1.5 × 10⁷ variants for CasMINI (an engineered Un1Cas12f1) and ~4.5 × 10⁷ variants for TnpB, targeting an average of ~2 amino acid changes per enzyme.
  2. Selection: Libraries were delivered into reporter cells via lentivirus. Cells were electroporated with guide RNAs (sgRNA for CasMINI, ωRNA for TnpB) and ssODN donors.
  3. Enrichment: EGFP-positive cells (indicating successful HDR) were isolated via Fluorescence-Activated Cell Sorting (FACS). Genomic DNA was extracted, and the nuclease genes were amplified to create the library for the next round.
  4. Iterative Rounds: The process involved four rounds of selection (two rounds of mutagenesis followed by purifying selection) to enrich for high-activity variants.
• Variant Assembly: Next-Generation Sequencing (NGS) identified enriched mutations. The authors combined these mutations into combinatorial variants (CVs), testing them against orthogonal targets to ensure generalizability.
• Validation: Top variants were tested at endogenous genomic loci in HEK293T cells, primary human T cells, and in vivo in mice (AAV delivery to the liver targeting PCSK9). Off-target effects were assessed using DISCOVER-Seq.

3. Key Contributions

• Cas12f1Super: A highly active variant of Un1Cas12f1 (derived from CasMINI) containing five specific mutations (K52E, K129E, E206K, K227E, K506E).
• TnpBSuper: A highly active variant of ISDra2 TnpB containing nine mutations (including E73K, L172V, V192L, etc.) combined with a codon-optimized sequence and a shortened ωRNA scaffold (Trim2).
• Mammalian-Directed Evolution Strategy: Demonstrated that selecting for HDR activity directly in human cells yields variants that outperform those evolved in bacteria or selected via single-point DMS, specifically by capturing synergistic multi-mutation effects.
• Base Editing Application: Showed that the evolved nuclease scaffold can be converted into a compact adenine base editor (dCas12f1Super) with superior performance.

4. Key Results

• Enhanced HDR Efficiency:
  • Cas12f1Super showed up to an 11-fold increase in HDR efficiency compared to the parental CasMINI in reporter assays.
  • TnpBSuper showed up to a 2.4-fold increase in HDR efficiency compared to wild-type TnpB.
  • When tested as a base editor, dCas12f1Super achieved up to a 10-fold improvement in editing efficiency over dCasMINI.
• Endogenous Editing:
  • Both super-variants showed significant improvements in HDR and Non-Homologous End Joining (NHEJ) across multiple endogenous loci in HEK293T cells.
  • In primary human T cells, TnpBSuper achieved up to a 6.9-fold increase in HDR efficiency.
  • At specific loci (e.g., TTR), TnpBSuper achieved editing efficiencies comparable to or slightly exceeding SpCas9, despite being much smaller.
• Specificity: DISCOVER-Seq analysis confirmed that the evolved variants retained the high specificity of their parental nucleases, with no increase in off-target effects despite the massive boost in on-target activity.
• In Vivo Efficacy (AAV Delivery):
  • Using AAV9 delivery in mice to target the PCSK9 gene (a therapeutic target for lowering cholesterol), both Cas12f1Super and TnpBSuper achieved high editing efficiencies in the liver (51% and 61%, respectively).
  • This resulted in a rapid and significant reduction in circulating PCSK9 protein levels, demonstrating therapeutic potential.
• Mechanism: In vitro cleavage assays showed that the improved activity was not due to increased catalytic cleavage rates in a test tube. This suggests the improvements stem from factors specific to the mammalian cellular environment, such as better protein stability, expression, or interaction with host DNA repair machinery.

5. Significance

This study represents a major breakthrough in the development of compact genome editors suitable for clinical translation.

1. Therapeutic Viability: By overcoming the size constraints of AAV vectors while simultaneously solving the low-activity problem of miniature nucleases, these tools (Cas12f1Super and TnpBSuper) make in vivo gene editing for a wider range of genetic diseases feasible.
2. Methodological Advance: The paper validates that directed evolution in mammalian cells is a superior strategy for optimizing genome editors compared to bacterial systems, as it captures complex host-specific interactions.
3. Versatility: The success of these variants in NHEJ, HDR, and base editing applications highlights their utility as a versatile platform for diverse therapeutic strategies, including precise gene correction and gene disruption.
4. Safety: The retention of high specificity ensures that the enhanced potency does not come at the cost of increased genomic toxicity, a critical requirement for clinical applications.