Delivering on the promise of gene editing

As gene editing technologies like CRISPR progress toward clinical study, researchers must continue to advance new approaches and address inherent challenges, explains Jon Chesnut, PhD, Senior Director, Cell Biology R&D, Thermo Fisher Scientific.

Gene editing tools such as zinc finger nucleases, transcription activator-like effector nucleases (TALEN) and clustered regularly interspaced short palindromic repeat (CRISPR) nucleases have been heralded for their enormous potential to treat diseases and genetic disorders. These platforms all induce double strand cuts in the chromosomal DNA that can be sealed by the cell, leading to a specific gene disruption, or resulting in a new target site for inserting DNA segments. Now, after nearly 35 years of research and more than a decade of preclinical progress, several different gene editing modalities are being tested in early phase clinical trials. Delivering on the therapeutic promise of these technologies relies not only on the results of these trials, but also on additional innovation aimed at addressing the limitations of current strategies and protocols. 

Early phase clinical trials for gene editing therapies

Numerous early phase clinical trials of potential gene editing therapies have emerged over the past five years. The majority are employing CRISPR (>120) with nearly 20 trials using the other nucleases mentioned above. These tools show promise to treat a wide range of diseases such as cancer (various leukaemias, multiple myeloma), blood (haemophilia, sickle cell), and metabolic disorders (type I diabetes). 

Gene editing challenges and potential solutions

As clinical trials that are testing the safety and efficacy of various gene editing therapies continue to progress, it is important to recognise the limitations of today’s tools and advance solutions that minimise the risk of any adverse effects. For example, it is becoming increasingly critical to:

1. Fine tune specificity. Gene editing tools can sometimes bind to unintended sites, typically because of sequence homologies and/or mismatch tolerance. Inducing double strand breaks at unintended locations in the genome can lead to off-target genetic modifications that are difficult to predict and potentially deleterious. Complicating matters even more, researchers currently lack effective and reliable methods for detecting and measuring off-target genomic events.

Potential solutions: For CRISPR systems, the likelihood of off-target editing has been shown to be influenced by a variety of factors, ranging from the design of the gRNA¹ and Cas9 protein², to the targeted cell type³, to name a few examples. The industry is addressing these challenges and improving both the specificity and sensitivity of gene editing tools by working toward creating new versions of identifying relatives to Cas9 that have higher specificity, activity, or broader target ranges. Ideally, improvements to CRISPR and other platforms would achieve all three.

2. Reduce/eliminate chromosomal disruptions caused by double-strand DNA breaks. Standard CRISPR systems generate double-strand breaks in the DNA which, at times, can introduce unintended chromosomal damage ranging from small insertions or deletions (indels) to large scale chromosomal deletion and translocation events. 

Possible solutions: A new generation of gene editing tools is being developed to avoid double stranded breaks altogether and to be potentially less destructive than traditional nuclease systems. Base editing involves using a mutated, inactive Cas9 nickase fused to a deaminase that drives specific and permanent single nucleotide changes via a single DNA strand nick. Although base editing has been shown to have significantly reduced off target damage than standard CRISPR systems, it can only produce four single-nucleotide swaps (C>T, G>A, T>C, and A>G)⁴. Prime editing, while also employing a single strand nick, expands this scope to all 12 possible changes using a more complex prime editing guide RNA (pegRNA). Prime editing can be used for base swapping, as well as for targeted small insertions and deletions. In one example, Choi et al. used a prime editing-based method called PRIME-Del that induces a deletion using a pair of pegRNAs  targeting opposite DNA strands, effectively programming not only the nicked sites but also the repair outcome⁵.

3. Improve delivery: Clinical translation of gene therapy is dependent on the efficient and safe delivery of gene editing tools into cells, and this can occur either ex vivo or in vivo. For ex vivo gene therapies, the genomes of particular cells are modified outside of the body, and then those cells are transplanted back into the patient. In vivo approaches involve delivering the gene editing components directly into the patient. Common delivery methods for gene editing tools are viral vectors, including adenoviral vectors (AdVs), recombinant adeno-associated viruses (rAAVs), and lentivirus vectors (LVs). Each comes with its own drawbacks, including immunogenicity, size restrictions, and off-target effects⁶. While viral vectors can be highly active, they can lead to only transient expression (rAAV, AdV), can lead to adverse immune responses (AdV) or can integrate randomly into the genome (LV). While LV can create sustained expression of the genome editing tool, the randomness of integration could lead to activation of oncogenes and poor regulation of the genome editing tool expression. In all cases, viral delivery vectors have limited payload capacity which can limit various use cases.

Potential solutions: Because of safety concerns regarding immunogenicity and insertional mutagenesis, as well as other challenges associated with viral-based donors, researchers are pursuing additional methods for ex vivo and in vivo delivery. Topping the list of alternatives are lipid nanoparticles (LNPs) and electroporation. In vivo delivery of LNP-packaged gene editing tools has been demonstrated successfully in animal models⁷ and we are currently developing LNP formulations that target specific organs in humans. Electroporation involves utilising high-voltage currents to permeabilise cell membranes so that gene editing tools can be delivered directly into cells. This method is used to modify genes in T cells⁸ and hematopoietic stem and progenitor cells (HSPCs⁾⁹. Ultimately, the different strategies for delivery must all balance efficiency with efficacy, cost, and toxicity.

4. Expand representation across global populations. The standard reference genome used to design targets for gene editing experiments is mostly representative of the DNA of any individual, but it does not account for the genetic diversity present in global populations. Researchers need to better understand these natural genetic variations and be mindful of them when designing gene editing therapeutics or even research experiments.

For example, Scott and Zhang¹⁰ looked at how human genetic variation impacts Cas endonuclease target choice, concluding that 21-35% of the targets they analysed contained PAM-altering variants. These findings suggest that a significant portion of guide RNAs designed to target these regions will have low efficacy. Notably, Scott and Zhang found ample targets for consideration in most genomic regions. Likewise, Lessard et al.¹¹ found that human genetic variation can alter the off-target landscape genome-wide, underscoring the importance of considering individual genomes when designing and evaluating CRISPR-based therapies to minimise risk of treatment failure and/or adverse outcomes. 

Possible solutions: Adding diversity to the reference database and applying advanced computational analysis will help expand representation, and there have been efforts over the past few years to sequence individuals from various ethnic groups around the world. We are currently curating data from the Genome Aggregation Database (gnomAD) and integrating it with TrueDesign, our free online tool that enables researchers to design, select, and order reagents for genome editing experiments. For CRISPR-Cas9 systems, the consideration of genetic variation can influence the choice of Cas endonuclease and perturbations, and the availability of multiple Cas enzymes with varying PAM requirements offers researchers a wide range of options.    

Recent advances in gene editing technologies are extremely encouraging, with simple gene modifications becoming routine in animal models. Researchers are identifying and engineering nucleases to improve specificity and developing new techniques such as base editing and prime editing to mitigate off-target chromosomal damage. In the future, the industry can direct more direction towards gene editing therapies that focus on gene expression modulation through activation and inhibition. These therapies could have even broader and more valuable impacts for the treatment of diseases caused by genetic abnormalities. Research across gene editing is changing medicine and it is invigorating to be here now, realising the tremendous therapeutic potential of these revolutionary technologies. 

References

1. CRISPRscan: designing highly efficient sgRNAs for CRISPR/Cas9 targeting in vivo 
2. DNA targeting specificity of RNA-guided Cas9 nucleases 
3. Whole-Genome Sequencing Analysis Reveals High Specificity of CRISPR/Cas9 and TALEN-Based Genome Editing in Human iPSCs
4. Single-nucleotide editing: from principle, optimization to application
5. Precise genomic deletions using paired prime editing
6. Viral Vectors for the in Vivo Delivery of CRISPR Components: Advances and Challenges
7. LNP-mediated delivery of CRISPR RNP for wide-spread in vivo genome editing in mouse cornea
8. CRISPR-engineered T cells in patients with refractory cancer
9. Priming Human Repopulating Hematopoietic Stem and Progenitor Cells for Cas9/sgRNA Gene Targeting
10. Implications of human genetic variation in CRISPR-based therapeutic genome editing
11. Human genetic variation alters CRISPR-Cas9 on- and off-targeting specificity at therapeutically implicated loci