Broken String Technology Identifies Off-Target Gene Edits 

 

CRISPR-Cas gene editing has the potential to revolutionize therapeutics. However, one significant hurdle stands in the way: off-target edits. This wouldn’t be such a serious problem, provided labs had better tools to identify double-stranded DNA breaks.  

To make that a reality, Cambridge, U.K.-based Broken String Biosciences has developed a new technology that rapidly identifies and maps DNA breaks, giving companies critical information to drive more informed decision-making when developing gene editing therapies. 

“Off-target edits have been an enormous problem for any potential gene editing therapy, but the challenge has been measuring them,” said Wouter Meuleman, Partner at Illumina Ventures. “Broken String gives labs a simple break detection workflow that provides clear, unambiguous results.” 

 

Addressing Off-Target Edits 

Broken String co-founder Felix Dobbs, Ph.D., tackled the problem while working on his doctorate at Cardiff University. The answer was an NGS-based system that spawned a paper in Nature Communications and eventually the new company. 

Off-target edits are major concerns because they could adversely affect patient health – in extreme cases, even driving cancer development. Such outcomes would be disastrous for the therapeutic company and the industry as a whole. Many early viral vector-based gene therapies were hamstrung because they could cause adverse events. Gene editing companies are determined not to repeat that history. 

“We are on the brink of a new horizon and must ensure we don't end up in a situation where these therapies can't reach patients because there are safety problems,” said Dobbs. “Gene editing has the potential to be absolutely transformative, but there are significant, knowable risks. Broken String wants to give biopharma companies the tools to assess those risks and develop the safest possible therapies.” 

 

INDUCE-seq and Advanced Therapeutic Development  

Broken String’s technology, INDUCE-seq, works with existing NGS platforms. It analyzes cells with double-strand DNA breaks, cross-links them to prevent the damage from being repaired and attaches adapters in situ in the cell nuclei to label breaks. Using a PCR-free approach, the DNA is extracted and given a second adapter. As a result, only DNA segments with breaks are sequenced, revealing their position and frequency in the genome. 

This technology comes at a pivotal time for gene editing companies. The U.S. Food & Drug Administration has been gradually increasing pressure on the industry to quantify off-target risks for their investigational new drug (IND) applications.  

“We are focused on supporting customers towards the end of the lead selection, going into preclinical IND work when they need this critical, supporting information,” said Dobbs. “They have their lead candidates, but now they must demonstrate their therapy won’t introduce significant, off-target edits in patient genomes.” 

Broken String’s technology helps companies produce the necessary documentation to support their regulatory filings. In fact, with no other break detection products on the market, Broken String could be the first to set the standard for off-target edit measurement. 

DNA double-strand break data could also refine therapeutic development. Current approaches lean heavily on legacy, small-molecule approaches. In the early stages, companies iteratively test hundreds of guides, primarily focusing on efficacy. Later, as they narrow down their options, safety considerations become more relevant.  

However, with INDUCE-seq providing more accessible information on CRISPR targeting, safety could be brought into the equation much earlier, de-risking assets upfront. Companies would have safer products and less risk of getting burned as their designs advance toward an IND. 

Even at this early stage, Broken String is seeing high demand for INDUCE-seq and is working with early adopters to integrate the technology into their workflows. The company has partnered with various biopharma companies, large and small, and anticipates significant commercialization in early 2025.  

Looking beyond that, Broken String plans to incorporate deep learning and other advanced computational techniques into the product to enhance off-target edit discovery and better predict their ramifications.   

“Ultimately, we should be able to predict how likely it will be for any given site on the genome to break,” said Dobbs. “We have early data from deep neural networks and other approaches using collated off-target data to predict where to edit safely. We could de-risk the process while also informing design to build systems that are optimized for both safety and efficiency.”