Researchers at New York University (NYU), Columbia University, and the New York Genome Center have developed an artificial intelligence (AI) platform that can predict on- and off-target activity of CRISPR tools that target RNA instead of DNA.
The team paired a deep learning model with CRISPR screens to control the expression of human genes in different ways, akin to either flicking a light switch to shut them off completely or by using a dimmer knob to partially turn down their activity. The resulting neural network, which they called targeted inhibition of gene expression via gRNA design— TIGER—was able to predict efficacy from guide sequence and context. The team suggests the new technology could pave the way to the development of precise gene controls for use in CRISPR-based therapies.
“Our deep learning model can tell us not only how to design a guide RNA that knocks down a transcript completely, but can also ‘tune’ it—for instance, having it produce only 70% of the transcript of a specific gene,” said Andrew Stirn, a PhD student at Columbia Engineering and the New York Genome Center. Stirn is co-first author of the researchers’ published paper in Nature Biotechnology, titled “Prediction of on-target and off-target activity of CRISPR-Cas13d guide RNAs using deep learning.” In their paper, the researchers concluded, “We believe that TIGER predictions will enable ranking and ultimately avoidance of undesired off-target binding sites and nuclease activation, and further spur the development of RNA-targeting therapeutics.”
CRISPR gene editing technology has many uses in biomedicine and beyond, from treating sickle cell anemia to engineering tastier mustard greens. While some CRISPR techniques target DNA using an enzyme called Cas9, in recent years, scientists have discovered another type of CRISPR that instead targets RNA using an enzyme called Cas13. “New CRISPR technologies hold great promise for a new generation of therapeutic agents,” the authors stated. “Among these, RNA-targeting CRISPR proteins have recently been shown to provide therapeutic values in disease models.”
RNA-targeting CRISPRs can be used in a wide range of applications, including RNA editing, knocking down RNA to block expression of a particular gene, and high-throughput screening to determine promising drug candidates. Researchers at NYU and the New York Genome Center previously created a platform for RNA-targeting CRISPR screens using Cas13 to better understand RNA regulation and to identify the function of noncoding RNAs. Because RNA is the main genetic material in viruses including SARS-CoV-2 and flu, RNA-targeting CRISPRs also hold promise for developing new methods to prevent or treat viral infections. Also, in human cells, when a gene is expressed, one of the first steps is the creation of RNA from the DNA in the genome.
“Similar to DNA-targeting CRISPRs such as Cas9, we anticipate that RNA-targeting CRISPRs such as Cas13 will have an outsized impact in molecular biology and biomedical applications in the coming years,” said Neville Sanjana, PhD, associate professor of biology at NYU, associate professor of neuroscience and physiology at NYU Grossman School of Medicine, a core faculty member at New York Genome Center, and the study’s co-senior author. “Accurate guide prediction and off-target identification will be of immense value for this newly developing field and therapeutics.”
A key goal of the team’s newly reported study was to maximize the activity of RNA-targeting CRISPRs on the intended target RNA—”on-target” activity—and minimize “off-target” activity on other RNAs, which could otherwise have detrimental side effects for the cell. “High precision is key to the safety of therapeutic RNA-targeting CRISPR agents,” they pointed out. Off-target activity includes both mismatches between the guide and target RNA as well as insertion and deletion mutations.
However, earlier studies of RNA-targeting CRISPRs focused primarily on on-target activity and mismatches, and predicting off-target activity, particularly insertion and deletion mutations, hasn’t been well-studied. In human populations, about one in five mutations are insertions or deletions, so these are important types of potential off targets to consider for CRISPR design.
For their study described in Nature Biotechnology, Sanjana and colleagues carried out a series of pooled RNA-targeting CRISPR screens in human cells. They measured the activity of 200,000 guide RNAs targeting essential genes in human cells, including both “perfect match” guide RNAs and off-target mismatches, insertions, and deletions. “In this study, we generated a large Cas13d dataset that measures the activity of ~200,000 gRNAs across multiple human cell lines and performed a comprehensive assessment of Cas13d gRNA on-target and off-target activity.” Sanjana’s lab teamed up with the lab of machine learning expert and co-senior author David Knowles, PhD, to engineer a deep learning model they named TIGER, which was trained on the data from the CRISPR screens.
Comparing the predictions generated by the deep learning model and laboratory tests in human cells, TIGER was able to predict both on-target and off-target activity, outperforming previous models developed for Cas13 on-target guide design and providing the first tool for predicting off-target activity of RNA-targeting CRISPRs. “Machine learning and deep learning are showing their strength in genomics because they can take advantage of the huge datasets that can now be generated by modern high-throughput experiments,” said Knowles, assistant professor of computer science and systems biology at Columbia University’s School of Engineering and Applied Science and a core faculty member at New York Genome Center. “Importantly, we were also able to use ‘interpretable machine learning’ to understand why the model predicts that a specific guide will work well.”
Added co-first author Hans-Hermann Wessels, PhD, a senior scientist at the New York Genome Center, “Our earlier research demonstrated how to design Cas13 guides that can knock down a particular RNA. With TIGER, we can now design Cas13 guides that strike a balance between on-target knockdown and avoiding off-target activity.” Wessels was previously a postdoctoral fellow in Sanjana’s laboratory.
By combining artificial intelligence with an RNA-targeting CRISPR screen, the researchers envision that TIGER’s predictions will help avoid undesired off-target CRISPR activity and further spur development of a new generation of RNA-targeting therapies.
The researchers also demonstrated that TIGER’s off-target predictions can be used to precisely modulate gene dosage—the amount of a particular gene that is expressed—by enabling partial inhibition of gene expression in cells with mismatch guides. This may be useful for diseases in which there are too many copies of a gene, such as Down syndrome, certain forms of schizophrenia, Charcot-Marie-Tooth disease (a hereditary nerve disorder), or in cancers where aberrant gene expression can lead to uncontrolled tumor growth. “… our model can be used for precise modulation of target RNA knockdown at scale,” they noted. “Specifically, our study suggests that RNA-targeting CRISPR perturbations can be used to systematically study the effect of gene dosage at the RNA levels,” they said.
“As we collect larger datasets from CRISPR screens, the opportunities to apply sophisticated machine learning models are growingly rapid,” Sanjana said. “We are lucky to have David’s lab next door to ours to facilitate this wonderful, cross-disciplinary collaboration. And, with TIGER, we can predict off-targets and precisely modulate gene dosage which enables many exciting new applications for RNA-targeting CRISPRs for biomedicine.”
This latest study further advances the broad applicability of RNA-targeting CRISPRs for human genetics and drug discovery, building on the NYU team’s prior work to develop guide RNA design rules, target RNAs in diverse organisms including viruses like SARS-CoV-2, engineer protein and RNA therapeutics, and leverage single-cell biology to reveal synergistic drug combinations for leukemia.
In their paper, the authors concluded, “Taken together, we believe that the ability to model the effect of nucleotide mismatches not only allows for an enhanced understanding of gRNA on-target specificity and off-target avoidance but also enables precise target knockdown to a defined degree that will be useful for diverse transcriptome engineering applications.”