TFU72 is a novel and potent DNA-PKcs inhibitor for enhancing homology-directed repair gene editing

This paper introduces TFU72, a novel and potent DNA-PKcs inhibitor that significantly enhances homology-directed repair gene editing efficiency by up to 30-fold while providing strategies to mitigate associated genotoxic risks for safe therapeutic and research applications.

The Gist

Imagine your DNA is a massive, intricate library containing the instruction manuals for building and running a human being. Sometimes, a typo (a genetic disease) appears in one of these books. Gene editing is like sending a team of expert editors into the library to find that specific page, cut out the typo, and paste in the correct text.

However, there's a problem: The library has a very aggressive "security guard" called NHEJ. When the editors make a cut, this guard rushes in immediately to tape the pages back together. But the guard is clumsy; it often tapes them back with a gap, a crumple, or a random note stuck in the middle. This creates a new error called an INDEL (Insertion or Deletion), which can make the book unreadable.

The editors want to use a "Homology-Directed Repair" (HDR) method instead. This is a slow, careful process where they use a perfect "donor template" (a clean copy of the page) to rewrite the text exactly right. But the clumsy security guard (NHEJ) is so fast and strong that it usually beats the careful editors to the punch.

Enter the New Hero: TFU72

This paper introduces a new tool called TFU72. Think of TFU72 as a high-tech "pause button" for that clumsy security guard.

1. The Problem with the Old Pause Button: Scientists previously used a tool called AZD7648 to pause the guard. It worked well, but it was a bit "sloppy." It didn't just pause the guard; it also accidentally tripped up other important library staff (other enzymes like ATM and ATR) who were actually helping the editors do their job. This caused side effects, like accidentally damaging other books in the library (off-target mutations) or tearing out huge chunks of pages (large deletions).
2. The New, Precise Pause Button: TFU72 is the new, upgraded version. It is like a laser-guided remote control. It hits the clumsy security guard (DNA-PKcs) with extreme precision and stops it dead in its tracks, but it leaves the helpful staff (ATM/ATR) completely alone.
   • The Result: Because the guard is stopped, the careful editors (HDR) have a wide-open window to do their job. In the lab, this new tool boosted the success rate of fixing genes by up to 30 times compared to doing nothing.

Testing the Tool in Real Life

The researchers didn't just test this on simple paper models; they tested it on "real" human cells that could be used for actual cures:

• Stem Cells (The Builders): Cells that can turn into blood.
• T-Cells (The Soldiers): Immune cells used in cancer therapies.
• iPSCs (The Chameleons): Cells that can become any type of tissue.

In all these cases, TFU72 worked like a charm, helping the editors insert new genetic instructions with high efficiency.

The Safety Catch: Avoiding Accidents

Even with a laser-guided remote, stopping the security guard for too long can cause chaos. If you leave the guard paused for 24 hours, the library gets messy:

• Off-target mutations: The editors might accidentally cut the wrong book because they are working too long without supervision.
• Chromosomal Translocations: Two different books might get glued together by mistake.
• Large Deletions: Huge chunks of a book might go missing.

The paper found a clever way to fix this: Timing is everything.

• The Analogy: Imagine you are fixing a car while the mechanic is on a coffee break. If you leave the car open for 24 hours, dust gets in, and parts might get lost. But if you fix it quickly and close the hood before the mechanic comes back, everything stays safe.
• The Solution: The researchers found that using TFU72 for a short burst (12 hours) instead of a long day (24 hours) gave the editors enough time to do their work but stopped the "messy" side effects.
• The Double Safety: They also combined TFU72 with a super-precise pair of scissors (HiFi Cas9). Using the precise scissors plus the short pause button meant the risk of accidental damage dropped to near zero.

The Bottom Line

This paper presents TFU72 as a game-changer for gene therapy. It's a smarter, more precise way to pause the body's natural "clumsy repair" system, allowing doctors to fix genetic diseases much more effectively.

By combining this new tool with precise editing scissors and short treatment times, scientists can now edit human cells with high success rates while keeping the risk of accidental damage very low. This brings us one big step closer to safe, reliable cures for genetic diseases like sickle cell anemia and other conditions.

Technical Summary

1. Problem Statement

Precise genome editing via Homology-Directed Repair (HDR) is critical for therapeutic applications but is often inefficient due to competition from the error-prone Non-Homologous End Joining (NHEJ) pathway.

• Current Limitations: While inhibiting DNA-dependent protein kinase catalytic subunit (DNA-PKcs) has been shown to enhance HDR, existing inhibitors like AZD7648 suffer from off-target kinase inhibition (specifically ATM and ATR, which are crucial for HDR) and induce significant genotoxicity.
• Genotoxic Risks: Previous studies using DNA-PKcs inhibitors have reported increased frequencies of off-target mutations, chromosomal translocations, and large on-target deletions, raising safety concerns for clinical translation.
• Need: There is a need for a novel, highly potent, and selective DNA-PKcs inhibitor that maximizes HDR efficiency while minimizing unintended genotoxic outcomes in both cell lines and therapeutically relevant primary human cells.

2. Methodology

The study employed a comprehensive approach to characterize a novel compound, TFU72, and compare it against the benchmark AZD7648.

• Compound Characterization:
  • Biochemical Assays: Determined IC50 values for DNA-PKcs and other PI3K family members (PI3Kα, β, γ, δ) and related kinases (ATM, ATR, mTOR).
  • Kinome Profiling: Screened TFU72 and AZD7648 against a panel of 513 human kinases to assess off-target inhibition profiles.
• Gene Editing Platforms:
  • Cell Lines: K562, HEK293, and U2OS cells were edited using Cas9 RNP and ssODN donors.
  • Primary Human Cells: Human iPSCs, CD34+ Hematopoietic Stem and Progenitor Cells (HSPCs), and T cells were edited using HiFi Cas9 (or WT Cas9 for comparison) RNP and AAV6-based HDR donor templates.
  • Targets: Multiple loci were targeted, including AAVS1, BCL11A, FANCF, HBB, CCR5, and STING1.
• Optimization Strategies:
  • Dose Titration: Tested concentrations from 0.25 μM to 4 μM.
  • Incubation Time: Evaluated the impact of compound incubation duration (4h, 8h, 12h, 24h, 72h) and delayed start times on HDR efficiency and genotoxicity.
  • Cas9 Variants: Compared HiFi Cas9 vs. Wild-Type (WT) Cas9 to assess impact on off-target effects.
• Genotoxicity Assessment:
  • Off-target INDELs: Quantified via NGS at known off-target sites.
  • Translocations: Measured via ddPCR between on-target and off-target sites.
  • Large Deletions: Analyzed using Oxford Nanopore sequencing of large amplicons (up to 10kb).
  • Viability: Assessed via MTT assays, flow cytometry, and Colony Forming Unit (CFU) assays.

3. Key Contributions

• Discovery of TFU72: Identification of a novel DNA-PKcs inhibitor with superior potency (Cellular IC50: 53 nM) compared to AZD7648 (180 nM).
• Superior Selectivity: TFU72 demonstrates a distinct kinase profile, sparing ATM and ATR (IC50 >10,000 nM) which are critical for HDR, whereas AZD7648 inhibits ATM (IC50: 83 nM).
• Mitigation Protocol: The authors established a specific workflow to mitigate genotoxicity:
  1. Use of HiFi Cas9 (reduces off-targets/translocations >10-fold vs. WT).
  2. Reduction of TFU72 incubation time to 12 hours (mitigates off-target INDELs and large deletions).
• Broad Applicability: Demonstrated efficacy across diverse cell types (immortalized lines, iPSCs, HSPCs, T cells) and insert sizes (short ssODNs to multi-kb AAV6 inserts).

4. Key Results

• HDR Enhancement:

  • TFU72 enhanced HDR efficiency by up to 30-fold in various contexts.
  • In primary cells (iPSCs, HSPCs, T cells), TFU72 achieved HDR rates comparable to or exceeding AZD7648 (e.g., increasing T cell HDR from ~1% to ~80%).
  • It effectively enhanced HDR even with "seemingly inactive" gRNAs that produce low INDEL frequencies, suggesting it blocks rapid NHEJ repair of DSBs, allowing time for HDR engagement.
  • Efficacy was maintained for large knock-ins (1kb–2.2kb) even at low AAV6 multiplicities of infection (MOI).

• Selectivity and Toxicity:

  • Kinase Profile: TFU72 showed >98% inhibition of DNA-PKcs but minimal inhibition (<30%) of most other kinases, specifically sparing ATM and ATR.
  • Cell Viability: TFU72-treated cells showed equal or slightly improved viability compared to AZD7648-treated cells.
  • Genotoxicity:
    • Like AZD7648, TFU72 increased off-target INDELs, translocations, and large deletions when used with WT Cas9 for 24h.
    • Mitigation Success: Switching to HiFi Cas9 reduced off-target INDELs and translocations by >10-fold. Reducing the incubation time to 12 hours normalized off-target INDELs and large deletion frequencies to levels similar to untreated controls, without sacrificing HDR enhancement.

• Specificity of Outcomes:

  • In T cells, TFU72 treatment uniquely induced a small frequency of MMEJ (Microhomology-Mediated End Joining) alleles, which was not observed in other cell types or with AZD7648 alone.

5. Significance

• Therapeutic Safety: This study provides a validated, optimized protocol (HiFi Cas9 + 12h TFU72 incubation) that allows for high-efficiency HDR editing in primary human cells while significantly mitigating the genotoxic risks (off-targets, translocations, large deletions) historically associated with DNA-PKcs inhibition.
• Clinical Relevance: The ability to enhance HDR in HSPCs and T cells (key targets for ex vivo gene therapies) with a compound that has a favorable safety profile and does not require cell washing (in some contexts) makes TFU72 a valuable tool for developing next-generation gene therapies.
• Mechanistic Insight: The findings reinforce that blocking NHEJ creates a time window for HDR, but the duration of this block must be carefully balanced to prevent the accumulation of genomic instability.
• Tool Availability: TFU72 is made available as a commercial product (PURedit™ HDR Enhancer), facilitating broader adoption in research and therapeutic development.

In conclusion, TFU72 represents a significant advancement in the genome editing toolbox, offering a potent, selective, and safer method to drive precise gene integration for both research and clinical applications.