In just over a decade, clustered regularly interspaced short palindromic repeats (CRISPR)-CRISPR-associated (Cas) systems have advanced from first demonstration as a programmable RNA-guided DNA nuclease to use in human clinical trials to provide durable cures for previously incurable genetic diseases. CRISPR-Cas genome editing relies on a guide RNA (gRNA) to target a specific sequence in the genome prior to the generation of a double-strand DNA break (DSB). This mechanism can be exploited to disrupt, insert, or replace nucleotide sequences of interest, providing a powerful method for modulating the human genome for therapeutic benefit. Despite CRISPR-Cas genome editing holding tremendous promise, clinical efficacy and safety concerns over making permanent changes to the human genome remain. This work describes multiple efforts to advance CRISPR-Cas genome editors for improved translation as a human therapeutic, with a focus on increased safety and precision.Two challenges for translating CRISPR-Cas systems for genome editing in humans are delivery and immunogenicity. While commonly used CRISPR effectors, such as Cas9, have shown great promise, their large size makes therapeutic delivery difficult. Moreover, pre-existing human adaptive immune responses to Cas9 may limit efficacy in vivo. To address these challenges, we engineered CasX, a novel and miniature CRISPR effector from non-human-associated microbes, which shows promise for improved delivery and reduced immunogenicity. To achieve this, we determined the cryo-electron microscopy (cryo-EM) structure of a CasX ortholog and used these structural insights to rationally engineer multiple CasX proteins and its gRNA for improved human cell genome editing activity.
In addition to safety concerns regarding CRISPR-Cas genome editing at unintended sites in the genome, another concern is the precision of genomic outcomes at the intended target site. Using two orthogonal assays, we identified that single-arm and entire-chromosome loss was a result of Cas9 DSBs in primary human T cells. Cas9-induced chromosome loss was specific to the targeted chromosome but was a generalizable phenomenon across different gRNAs targeted throughout the genome. T cells with chromosome loss persisted for weeks during ex vivo cell culture but chromosome loss was surprisingly mitigated by simply changing the order of operations during the genome editing protocol. This unexpected finding may explain the lack of Cas9-induced chromosome loss we observed in T cells from patients in a first-in-human phase I clinical trial, and provides the first method for avoiding this potential genotoxicity.
Before ensuring that unintended genomic outcomes are avoided during genome editing, targeting of unintended cell types in vivo needs to be avoided. We developed virus-like particles (VLP) that package transient Cas9 ribonucleoproteins (RNP) with or without a transgene of interest. VLPs were capable of complex genome engineering, including both gene insertion and gene disruption, in primary immune cells to yield therapeutic chimeric antigen receptor (CAR) T cells. VLP-based manufacturing of CAR T cells can occur in a single step, greatly simplifying the standard manufacturing of these approved therapies. In addition, we showed that VLPs can be pseudotyped to enable cell-type specific genome editing of CD4+ T cells within a mixed cell population.
Finally, during the COVID-19 pandemic our ongoing efforts to improve the safety and efficacy of CRISPR-Cas genome editing were paused. However, in response to a shortage of clinical testing in our community, we swiftly pivoted our scientific skills and expertise toward building a SARS-CoV-2 clinical testing facility. We established organizational, safety, scientific, regulatory, and clinical practices to ultimately test thousands of patent samples and provide clinical COVID-19 diagnoses.