Anti-CRISPR-mediated continuous directed evolution of CRISPR-Cas9 in human cells

This study introduces CRISPR-MACE, a continuous evolution platform that utilizes anti-CRISPR-mediated selection within human cells to successfully engineer novel Cas9 variants with enhanced DNA binding capabilities and significantly improved resistance to inhibitors, overcoming the limitations of traditional bacterial-based optimization.

The Gist

The Big Idea: Evolving Super-Tools Inside Human Cells

Imagine you are trying to build a better version of a specific tool, like a Swiss Army knife. Usually, scientists build these tools in a simple workshop (bacteria) and then hope they work perfectly when they take them to a complex, messy construction site (human cells).

The Problem: Often, the tool works great in the simple workshop but gets jammed or breaks in the messy construction site. The environment is just too different.

The Solution: This paper introduces a new method called CRISPR-MACE. Instead of building the tool in a simple workshop and hoping for the best, they built a "living factory" inside the human cells themselves. This factory constantly tries to break the tool, fix it, and make it stronger, all while the tool is actually working in the human environment.

---

The Cast of Characters

To understand how this works, let's meet the players in this biological drama:

1. The Tool (Cas9): Think of Cas9 as a pair of molecular scissors. Scientists use it to cut or edit DNA. In this experiment, they used a "dead" version (dCas9) that can't cut but can still stick to DNA and act as a magnet to turn genes on.
2. The Villain (AcrIIA4): This is a natural "anti-tool" protein. Bacteria use Cas9 to fight viruses, and the viruses evolved AcrIIA4 to jam the scissors. It's like a lock that fits perfectly over the handle of your Swiss Army knife, making it useless.
3. The Factory (Adenovirus): The scientists hijacked a harmless virus (Adenovirus) to act as the factory. This virus is designed to replicate (make copies of itself) only if the "Tool" works correctly.
4. The Mutator (Error-Prone Polymerase): Inside the factory, there is a machine that copies the virus's DNA. But this machine is broken on purpose! It makes typos (mutations) constantly. This creates millions of slightly different versions of the "Tool" every time the virus replicates.

---

How the "Living Factory" Works (The Analogy)

Imagine a high-stakes game of Escape Room inside a human cell.

1. The Setup: The scientists put the "Tool" (Cas9) and the "Villain" (AcrIIA4) inside a human cell. The Villain is trying to lock the Tool so it can't work.
2. The Goal: The Tool needs to find a specific spot on the DNA and stick to it. If it sticks, it flips a switch that tells the virus factory: "Great job! Make more copies of yourself!" If it fails, the virus dies out.
3. The Pressure: The scientists add a special drug (pomalidomide) that acts like a dimmer switch for the Villain.
   • Start of the game: The Villain is very strong (high concentration). Only the very best, luckiest versions of the Tool can escape the Villain and flip the switch.
   • The Evolution: As the virus replicates, the "Mutator" machine creates random typos. Most typos make the Tool worse. But occasionally, a typo makes the Tool better at escaping the Villain.
   • The Selection: The viruses with the "better" Tools survive and multiply. The viruses with "worse" Tools die.
   • Ramping Up: As the game progresses, the scientists turn the dimmer switch down, making the Villain even stronger. Now, only the super-evolved Tools can survive.

The Results: What Did They Find?

After running this "Escape Room" for several rounds, they found some amazing results:

• The "Gatekeeper" Mutation: In two separate experiments, the very first mutation to appear was the same one (changing a Glycine to an Aspartic acid at position 12). It was like finding the master key that unlocked the door. This mutation didn't make the Tool resistant to the Villain immediately, but it made the Tool stick to DNA better. This gave the Tool a head start.
• The Teamwork Effect: Once the Tool had that "Gatekeeper" mutation, other mutations could join in. Some mutations helped the Tool ignore the Villain, while others helped it stick to DNA even tighter.
• The Ultimate Champion: They created a "Super-Tool" (a quintuple mutant) that was 1,000 times more resistant to the Villain than the original. It could ignore the lock completely while still sticking to the DNA perfectly.

Why This Matters

• No More Guessing: Before this, scientists had to guess how to fix these tools in a simple lab setting, hoping they would work in humans. Now, they can evolve the tools directly inside human cells, ensuring they are perfectly adapted to the real environment.
• Future Applications: This method isn't just for scissors (Cas9). It could be used to evolve other gene-editing tools, making them faster, more accurate, or resistant to drugs that might stop them.
• The "Cheating" Problem Solved: In previous methods, cells would sometimes "cheat" (stop working to save energy). This new system uses a virus that must work to survive, so cheating is impossible. The selection is incredibly strict.

The Bottom Line

This paper describes a breakthrough where scientists turned human cells into a biological evolution gym. They forced a gene-editing tool to fight against a natural inhibitor, constantly mutating and testing itself until it became a super-soldier. The result is a new generation of CRISPR tools that are stronger, smarter, and perfectly tuned for use in human medicine.

Technical Summary

1. The Problem

• Limitation of Current Engineering: Most CRISPR-Cas system optimization relies on directed evolution in bacterial hosts (e.g., E. coli). However, bacterial cellular environments differ significantly from mammalian cells in terms of genome size, chromatin structure, and molecular crowding. Consequently, variants optimized in bacteria often exhibit sub-optimal or inactive performance in human cells.
• Lack of Continuous Evolution in Mammals: Traditional step-wise directed evolution in human cells is difficult to scale and suffers from "cheating" (selection escape) because the cell itself is the unit of selection. There is a critical lack of robust platforms that enable continuous directed evolution (simultaneous mutagenesis, selection, and amplification) directly within the complex mammalian cellular milieu.

2. Methodology: CRISPR-MACE

The authors developed CRISPR-MACE (Mammalian cell-enabled Adenovirus-assisted Continuous Evolution), a platform adapted from the MACE system to evolve CRISPR-Cas9 directly in human cells.

• Viral Vector System:

  • The system uses a replication-deficient adenovirus encoding a fusion protein: dCas9–TAD (catalytically dead Cas9 fused to a VPR transcriptional activation domain).
  • The viral genome lacks two essential genes: AdProt (adenoviral protease) and AdPol (adenoviral DNA polymerase).
  • Host Cells: Engineered human cells (HEK293A derivatives) constitutively express a highly error-prone variant of AdPol (EpPol) and inducibly express AdProt.
  • Mechanism: The EpPol introduces mutations at a high rate ($3.7 \times 10^{-5}$) during viral replication. Only viral genomes encoding functional dCas9–TAD variants that can activate the promoter-less AdProt gene can replicate and produce new virions.

• Selection Circuit & Tunable Pressure:

  • Target: The dCas9–TAD must bind guide RNAs (gRNAs) targeting the region upstream of the AdProt gene to activate transcription.
  • Selection Pressure (Anti-CRISPR): To select for specific traits, the system incorporates AcrIIA4, a potent anti-CRISPR protein that inhibits SpCas9 by mimicking DNA and blocking PAM binding.
  • Tunability: To prevent the selection pressure from being too high (which would kill the virus immediately), the authors fused AcrIIA4 to a superdegron (csd). This fusion protein is degraded in a dose-dependent manner upon treatment with pomalidomide. By titrating pomalidomide, researchers can gradually increase selection pressure over the course of the evolution campaign.

• Evolution Workflow:

  1. Wild-type dCas9–TAD adenovirus is introduced into selector cells containing EpPol, AdProt, and AcrIIA4–csd.
  2. Pomalidomide is used to degrade AcrIIA4–csd, allowing initial replication.
  3. Over successive passages, pomalidomide levels are reduced, increasing AcrIIA4 levels and selection pressure.
  4. Only viral variants that escape AcrIIA4 inhibition while retaining DNA-binding capability survive and amplify.
  5. Viral DNA is sequenced (Nanopore) after each passage to track mutation frequencies.

3. Key Contributions

• First Continuous Evolution in Human Cells: Established a foundational platform for continuously evolving genome-targeting agents directly in human cells, bypassing the limitations of bacterial optimization.
• Tunable Selection Circuit: Developed a novel mechanism using pomalidomide-controlled degradation of anti-CRISPR proteins to precisely modulate selection pressure, enabling the stepwise evolution of complex traits.
• Discovery of Novel Cas9 Variants: Successfully evolved Streptococcus pyogenes Cas9 (SpCas9) variants with unprecedented properties, including high resistance to AcrIIA4 and altered DNA binding kinetics.

4. Key Results

• Evolutionary Trajectory:

  • Two independent evolution campaigns (polyclonal and monoclonal) converged on similar evolutionary paths.
  • Gatekeeper Mutation: The G12D mutation (Glycine to Aspartic acid at position 12) emerged first in both campaigns. While it only provided a modest (~2-fold) increase in AcrIIA4 resistance, it significantly enhanced DNA binding affinity (3-fold improvement in $K_d$). This suggests G12D acts as a "gatekeeper," improving DNA engagement to allow subsequent mutations to arise.
  • Synergistic Mutations: Later mutations (e.g., Q1221H, A764V, R919K, V955I) conferred strong AcrIIA4 resistance but often weakened DNA binding individually. However, when combined with the G12D background, the DNA binding deficit was compensated.
  • Final Variants: The quintuple mutant DVKIH (G12D/A764V/R919K/V955I/Q1221H) achieved a ~890-fold increase in IC50 against AcrIIA4 (nearly 1000-fold resistance) while maintaining robust DNA binding and transcriptional activity.

• Biochemical Characterization (BLI):

  • AcrIIA4 Resistance: The evolved variants showed dose-dependent resistance to AcrIIA4, with the quintuple mutant effectively preventing inhibitor-mediated disruption of DNA binding even at high inhibitor concentrations.
  • DNA Binding Kinetics: Variants exhibited diverse effects on DNA binding. Some (like G12D/Q1221H) showed a 10-fold increase in DNA affinity ($K_d \approx 0.4$ nM), while others (like R919K) showed reduced affinity. The final evolved variants balanced these opposing forces to maintain function.

• Cellular Validation:

  • Transcriptional Activity: Evolved variants retained or improved transcriptional activation (CRISPRa) in human cells, even in the presence of high AcrIIA4 levels.
  • CASFISH Imaging: Variants with enhanced DNA affinity (G12D and DH) were used in Cas9-mediated Fluorescence In Situ Hybridization (CASFISH). They showed 2-fold higher signal intensity at pericentromeric repeats compared to wild-type, demonstrating improved utility for genomic imaging.

5. Significance

• Bridging the Gap: This work proves that continuous directed evolution can be successfully conducted in human cells, yielding variants that are functionally superior in the native mammalian environment compared to those evolved in bacteria.
• Complex Trait Optimization: The study demonstrates the ability to evolve proteins for interdependent and opposing functions (increasing DNA binding while decreasing inhibitor binding). The discovery of the "gatekeeper" mutation (G12D) highlights how evolutionary history and epistasis shape the accessibility of functional landscapes.
• Therapeutic and Research Impact:
  • Resistance to Inhibitors: The generation of Cas9 variants resistant to AcrIIA4 is crucial for therapeutic applications where endogenous or exogenous anti-CRISPRs might limit efficacy.
  • Enhanced Tools: Variants with higher DNA residence time and affinity offer improved sensitivity for imaging and potentially more efficient genome editing in difficult-to-target loci.
  • Platform Scalability: The CRISPR-MACE platform is generalizable. By swapping the anti-CRISPR protein or the selection circuit, it can be used to evolve other genome-targeting agents (base editors, prime editors, epigenetic modifiers) directly in human cells.

In summary, CRISPR-MACE represents a paradigm shift in protein engineering, moving from bacterial test tubes to the complex human cellular environment to discover next-generation CRISPR tools with tailored, high-performance properties.