A single Oxford Nanopore flow cell can now run CRISPR gene editing and DNA sequencing simultaneously, allowing researchers to watch in real time exactly where the editing machinery cuts—and where it misses its target. This matters because CRISPR therapies, including the recently approved £2.2 million-per-patient Casgevy, carry a risk of off-target edits that could cause cancer. Current methods for detecting these errors are slow and incomplete, especially for patients whose genomes differ from the standard reference—and human genetic diversity is poorly captured in existing datasets. The problem is even worse for livestock and crops, whose genomes are ten times more variable than human genomes. If this technology works at scale, it could let clinicians profile a patient’s own genome for off-target risks before treatment begins. For drug developers, it could cut the cost of screening CRISPR therapies, potentially making treatments viable for more of the roughly 5,000 genetic diseases that currently have no therapy. The same approach could also speed regulatory approval for edited crops and livestock by providing clear, per-genome safety data.
View original technical description
Just months ago the first CRISPR-based gene editing therapy (Casgevy™) was licensed in the UK, USA, and EU to treat sickle cell and related anaemias but at a reported cost of $2.2m per patient. Many other CRISPR based treatments are under development and awaiting regulatory approval but a consistent concern is the consequences of off-target genome editing. These events could be harmful, even causing cancer, and finding such events or even side effects in clinical trials will cause expensive failures which push up the cost of the successful versions of these new therapeutics. Whilst the majority of off-target events are classified as guide RNA (gRNA) dependent, whereby the CRISPR complex binds to an off-target site with a similar sequence to the target site, less common gRNA-independent off-target events have also been identified. In 2022 Nork et al. showed that 22 years after the publication of the first human genome, long read nanopore sequencing was finally able to completely sequence and assemble the human genome revealing a previously hidden 8% of the genome, an additional 238 million bases. Each human genome has millions of single nucleotide polymorphisms compared to the reference, and some contain large chunks (millions of bases) of novel sequence. Indeed human genetic diversity has been shown to alter off-target outcomes of CRISPR therapeutic gene editing. Our ability to predict in-silico the impact of genetic diversity on CRISPR activity is currently limited due to the bias in existing human variation datasets towards the European population. Ideally the on- and off- target activity profile for a CRISPR-based therapeutic needs to be assessed on a per patient basis, prior to treatment delivery. Similar off-target concerns exist for genome editing of livestock and crops, which also require regulatory approval. These genomes are often 10x more variable between breeds and cultivars than the human genome, and many key crops such as Wheat, Potatoes and Brassicas are polyploids, requiring edits of all gene copies, or just specific sub-genome homologues. Our project combines two disruptive technologies that are revolutionising genetics: CRISPR genome editing and Oxford Nanopore Technologies (ONT) sequencing. We have recently discovered how to modify the conditions in an ONT flow cell such that CRISPR/Cas9 can function while the nanopores continue to sequence normally. This allows us to rapidly profile the targeting specificity of a given gRNA, different CRISPR enzyme derivatives, and on different genomes. The largest commercial market for our approach would be screening CRISPR based human therapeutics, reducing development costs and enabling more genetic diseases to have therapeutics - currently there are ~5,000 such untreated diseases with 200 million sufferers. In this project we will develop our technology further (wet lab and bioinformatics), and benchmark it to existing standards. Early discussions under NDA with a US biotech executive in this field has suggested to us that this technology would be of great interest. However, wide adoption of this technology and subsequent impact would necessitate substantial pilot data, as we outline in this proposal. Thus, we find our team at the happy coincidence of being able to generate academically interesting and publishable research alongside knowledge which we can transfer to industry to maximise the public good.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know