In a recent Q&A session with BioPharma BoardRoom, Gavin Kurgan, Bioinformatics Applications Development Manager at Integrated DNA Technologies (IDT), delves into the cutting-edge methodologies and technologies IDT employs to ensure the safety and accuracy of CRISPR edits. Kurgan highlights IDT's use of optimized GUIDE-seq methods, rhAmpSeq technology, and innovative high-fidelity enzymes like HiFi Cas9, as well as the company's efforts in navigating regulatory challenges and integrating AI for enhanced quality assurance. Through strategic partnerships and a commitment to continuous innovation, IDT is setting new standards in the CRISPR gene editing landscape.

What specific methodologies does IDT employ to ensure the safety and accuracy of CRISPR edits?

For evaluation of the safety and accuracy of CRISPR edits through IDT’s newly available services, we employ two major technologies and workflows. For our off-target nomination service, we offer an optimized GUIDE-seq method which incorporates our rhAmp technology, coupled with our in-house data analysis pipeline. This nomination process also includes several quality control checks, like quantifying the integration rate of the tag at the intended on-target site, as well as editing a positive control site to ensure there will be appropriate sensitivity to qualify results. 

For off-target verification, we offer our rhAmpSeqTM technology coupled to proprietary analysis algorithms for classification of off-target editing and translocations with sensitivity as low as 0.1% editing frequencies. These algorithms are an improved version of our rhAmpSeq CRISPR Analysis Tool first published in 2021, with the addition of multiple new algorithms for statistical classification of verified off-target sites and characterization of translocation events.

Can you provide examples of how IDT has successfully minimized off-target effects in gene editing therapeutics?

Of course! This type of work started in 2014 as we worked to commercialize a high-quality Cas9 enzyme that inherently had fewer off-targets than the plasmid expression systems that were common product offerings at the time. Fast-forward to 2018, we published the foundational research behind our high-fidelity Cas9 enzyme (HiFi Cas9) and released this protein as a product in our CRISPR portfolio. This is an SpCas9 variant with highly reduced off-target editing that maintains the efficiency of on-target editing. Since then, we have also worked to engineer more active versions of higher fidelity enzymes, like the Cas12a family of enzymes, which can be seen through the release of our AsCas12a Ultra and LbCas12a Ultra enzymes. We have additionally been working to provide new enhancers to improve the efficiency of desired DNA repair events, like homology directed repair, without increasing off-target editing. These products are some of our latest innovations which we are planning to launch commercially soon. 

In addition, we are continually working to release these products in a stage-appropriate format for clinical work. Through our partnership with Aldevron, we are able to offer GMP versions of many of the same high-fidelity enzymes and reagents we have innovated at IDT. IDT and Aldevron are operating companies of Danaher Corporation (NYSE:DHR). 

What are some of the main regulatory challenges IDT faces in quality assurance for CRISPR technology, and how does the company navigate them?

Most of the quality assurance challenges we face in the CRISPR space are largely due to the novelty of the whole system as a therapeutic modality. For instance, CRISPR is unique in the sense that mutations introduced into both gRNAs and DNA donors used for homology directed repair could have detrimental effects without proper quality control procedures. For gRNAs, mutations in the molecule could result in decreased activity, or even worse, a novel gRNA targeting new putative regions in the genome in the case of mutations within the spacer region of the molecule. For homology directed repair DNA donors, similar issues can arise, and mutations in a DNA donor can result in incorporation of unintended mutations after CRISPR gene editing in the genome of interest. To navigate this, we implement a number of critical quality control assays for our CRISPR-related oligonucleotides, spanning different analytical platforms (ESI-MS, etc.), all the way to direct sequencing of the molecules, while also continually innovating to create and characterize new analytical assays. 

Can you elaborate on the critical considerations for detecting, assessing, and addressing CRISPR safety and efficacy?

One of the most critical considerations for accurately assessing the safety of CRISPR editing are standards and process controls to describe the analytical sensitivity and specificity of a method. I think there is a goal we need to work toward as a scientific community to shape our communication about CRISPR systems along the following lines: “I am X% confident that this method will be able to detect Y% of all off-targets down to Z% editing”—which is a statement that requires a diverse set of highly-characterized genome editing standards.  

Several efforts are currently on-going to address this need from a scientific community perspective, like those led by the Genome Editing Consortium headed by the National Institute of Standards and Technology that we, as an organization, participate actively in. Furthermore, there are individual efforts like the standards we have created to benchmark our off-target nomination and validation technologies-turned-services that we are offering following this most recent partnership with SeQure Dx. While these standards and process controls are improved upon, it will be critical to strive to use multiple orthogonal assays to ensure that safety is being accurately assessed. 

How does IDT integrate advanced technologies, such as AI and machine learning, to enhance quality assurance and production efficiency in gene editing?

Technologies like AI and machine learning are continuing to play an important role in the genomics space, and gene editing is no exception. This can be readily seen by a number of excellent pieces of scientific work, like the different modeling efforts to predict gene editing outcomes a priori by separate scientific groups, like those at MIT, the Wellcome Sanger Institute, and more. Several of IDT’s current models, like our SpCas9 on-target model, already implement this sort of technology to attempt to make sure that gRNAs chosen for experiments have high on-target editing efficiency. These types of models also have a place in production processes for gene editing reagents, for example, identifying problematic motifs for oligo synthesis, providing quantitative estimation of different synthesis by-products, estimating the effects of any unintended oligo species, and more.