Efficient and safe therapeutic use of paired Cas9-nickases for primary hyperoxaluria type 1

Laura Torella, Julia Klermund, Martin Bilbao‐Arribas, Ibón Tamayo, Geoffroy Andrieux, Kay O. Chmielewski, África Vales, Cristina Olagüe, Daniel Moreno, Ivan Raimondi, Amaya Abad, Julen Torrens-Baile, Eduardo Salido, Maite Huarte, Mikel Hernáez, Melanie Boerries, Toni Cathomen, Nerea Zabaleta, Gloria González‐AseguinolazaEMBO 2024 · DOI

TL;DR

This paper presents a novel CRISPR-Cas9 gene editing strategy using paired Cas9 nickases delivered via AAV vectors to safely and effectively treat primary hyperoxaluria type 1 (PH1) in mice. This approach significantly reduces off-target edits and AAV integration, offering a promising path towards a permanent, one-time therapeutic intervention for this rare genetic liver disease.

What Problem Does This Paper Address?

Primary hyperoxaluria type 1 (PH1) is a severe genetic disorder where the liver produces too much oxalate, leading to kidney damage, stones, and ultimately kidney failure. Current treatments, like RNAi-based therapies, require lifelong administration and have limitations. Gene editing with CRISPR-Cas9 offers the potential for a permanent cure by disrupting the gene responsible for oxalate production (HAO1). However, the clinical application of CRISPR-Cas9 is hampered by concerns about off-target edits (unintended DNA changes elsewhere in the genome) and the integration of the AAV delivery vector into the host genome, which could lead to unforeseen consequences. This study aims to address these safety and efficacy concerns by employing a modified CRISPR-Cas9 system.

Methods

The core innovation in this study is the use of paired Staphylococcus aureus Cas9 nickases (D10ASaCas9). Unlike standard Cas9 nucleases that create double-strand breaks (DSBs), nickases create single-strand breaks (nicks). By using two nickases with carefully spaced guide RNAs, the researchers induce targeted DSBs that are then repaired by cellular mechanisms. This "paired nickase" approach is theorized to be more precise, as it requires two specific binding events to occur in close proximity, significantly reducing the likelihood of off-target nicks.

These paired nickases were delivered into the liver cells of PH1 mice using adeno-associated viral (AAV) vectors. To assess the safety and specificity of this approach, the researchers employed several advanced techniques:

  • CIRCLE-seq and CAST-Seq: These methods are used to detect off-target cleavage events and chromosomal translocations, respectively, across the entire genome.
  • Amplicon sequencing (including a custom pipeline): This was used to analyze the precise genomic modifications at the target site and to quantify AAV integration events. The development of a custom pipeline was necessary to accurately characterize the complex insertions and deletions resulting from the paired nickase activity.
  • Western Blot and Immunohistochemistry (IHC): These techniques were used to measure the reduction in glycolate oxidase (GO) protein levels, the direct therapeutic target.

Datasets

Not discussed in the paper.

Key Insights and Ideas

  • Paired nickases are safer and more specific: The study strongly suggests that using paired Cas9 nickases significantly reduces off-target edits and chromosomal translocations compared to standard Cas9 nucleases. This is a critical step towards making CRISPR-based therapies safer for clinical use.
  • Reduced AAV integration: A surprising and highly significant finding was that paired D10ASaCas9 led to a substantially lower frequency of AAV vector integration at the target site compared to the nuclease version. This is crucial because AAV integration can disrupt gene function or lead to insertional mutagenesis.
  • Microhomology-Mediated End Joining (MMEJ) plays a key role: The observed heterogeneous on-target modifications, particularly deletions, were found to be repaired via MMEJ, a pathway that utilizes short homologous sequences. This repair mechanism appears to be less permissive to AAV integration.
  • All-in-one AAV vector simplifies delivery: The development of a single AAV vector carrying both the D10ASaCas9 and the two guide RNAs simplifies the delivery system and maintains high editing efficacy even at lower doses, paving the way for more practical therapeutic applications.
  • Limitations of current analytical tools: The study highlights that existing tools for analyzing CRISPR editing outcomes, especially for complex insertions and deletions from nickase activity, are not always sufficient, necessitating the development of custom bioinformatics pipelines.

Key Results

  • Effective HAO1 gene disruption: Paired D10ASaCas9 successfully disrupted the HAO1 gene in PH1 mice, leading to a significant decrease in GO protein expression.
  • Therapeutic effect in PH1 mice: Treated mice showed a reduction in oxalate accumulation and prevention of kidney damage, demonstrating a clear therapeutic benefit.
  • No detectable off-target activity or translocations: CIRCLE-seq and CAST-Seq analyses did not reveal any significant off-target edits or chromosomal translocations.
  • Significantly reduced AAV integration: AAV integration at the target site was reduced by approximately 5-fold when using paired D10ASaCas9 compared to wtSaCas9.
  • High indel frequency with low AAV integration: The all-in-one vector achieved high indel frequencies (up to 70%) with minimal AAV integration (as low as 0.03%) at the lowest tested dose.

Limitations and Future Work

  • Mouse model limitations: While promising, the study was conducted in a mouse model. Further validation in humanized models or preclinical studies is necessary to confirm safety and efficacy in humans, as AAV tropism and immune responses can differ between species.
  • Analytical tool development: The authors emphasize the need for continued development of robust analytical tools to accurately characterize complex genomic modifications and AAV integration events, especially for paired nickase systems.
  • Long-term durability and safety: While the study suggests a permanent reduction in GO expression, long-term studies are needed to confirm the durability of the therapeutic effect and to further assess any potential late-onset adverse events.
  • Delivery vector optimization for humans: The study used AAV8, which has higher transduction efficiency in mice than in human hepatocytes. Future work should focus on exploring novel AAV serotypes or alternative delivery methods (like nanoparticles) for efficient delivery to human liver cells.
  • Consideration of self-inactivating Cas9: The authors suggest exploring self-inactivating Cas9 variants to further enhance safety by limiting the duration of Cas9 expression.

Summary updated

Abstract

The therapeutic use of adeno-associated viral vector (AAV)-mediated gene disruption using CRISPR-Cas9 is limited by potential off-target modifications and the risk of uncontrolled integration of vector genomes into CRISPR-mediated double-strand breaks. To address these concerns, we explored the use of AAV-delivered paired Staphylococcus aureus nickases (D10ASaCas9) to target the Hao1 gene for the treatment of primary hyperoxaluria type 1 (PH1). Our study demonstrated effective Hao1 gene disruption, a significant decrease in glycolate oxidase expression, and a therapeutic effect in PH1 mice. The assessment of undesired genetic modifications through CIRCLE-seq and CAST-Seq analyses revealed neither off-target activity nor chromosomal translocations. Importantly, the use of paired-D10ASaCas9 resulted in a significant reduction in AAV integration at the target site compared to SaCas9 nuclease. In addition, our study highlights the limitations of current analytical tools in characterizing modifications introduced by paired D10ASaCas9, necessitating the development of a custom pipeline for more accurate characterization. These results describe a positive advance towards a safe and effective potential long-term treatment for PH1 patients.
BibTeX
@inproceedings{torella2024efficient,
  author = {Laura Torella and Julia Klermund and Martin Bilbao‐Arribas and Ibón Tamayo and Geoffroy Andrieux and Kay O. Chmielewski and África Vales and Cristina Olagüe and Daniel Moreno and Ivan Raimondi and Amaya Abad and Julen Torrens-Baile and Eduardo Salido and Maite Huarte and Mikel Hernáez and Melanie Boerries and Toni Cathomen and Nerea Zabaleta and Gloria González‐Aseguinolaza},
  title = {{Efficient and safe therapeutic use of paired Cas9-nickases for primary hyperoxaluria type 1}},
  booktitle = {EMBO},
  year = {2024},
  doi = {10.1038/s44321-023-00008-8},
}
Figure