Adoptive cellular therapy using chimeric antigen receptor (CAR) T cell therapies have produced significant objective responses in patients with CD19+ hematological malignancies, including durable complete responses. Although the majority of clinical trials to date have used autologous patient cells as the starting material to generate CAR T cells, this strategy poses significant manufacturing challenges and, for some patients, may not be feasible because of their advanced disease state or difficulty with manufacturing suitable numbers of CAR T cells. Alternatively, T cells from a healthy donor can be used to produce an allogeneic CAR T therapy, provided the cells are rendered incapable of eliciting graft versus host disease (GvHD). One approach to the production of these cells is gene editing to eliminate expression of the endogenous T cell receptor (TCR). Here we report a streamlined strategy for generating allogeneic CAR T cells by targeting the insertion of a CAR transgene directly into the native TCR locus using an engineered homing endonuclease and an AAV donor template. We demonstrate that anti-CD19 CAR T cells produced in this manner do not express the endogenous TCR, exhibit potent effector functions in vitro, and mediate clearance of CD19+ tumors in an in vivo mouse model.

Clinical translation of in vivo genome editing to treat human genetic diseases requires thorough preclinical studies in relevant animal models to assess safety and efficacy. A promising approach to treat hypercholesterolemia is inactivating the secreted protein PCSK9, an antagonist of the LDL receptor. Here we show that single infusions in six non-human primates of adeno-associated virus vector expressing an engineered meganuclease targeting PCSK9 results in dose-dependent disruption of PCSK9 in liver, as well as a stable reduction in circulating PCSK9 and serum cholesterol. Animals experienced transient, asymptomatic elevations of serum transaminases owing to the formation of T cells against the transgene product. Vector DNA and meganuclease expression declined rapidly, leaving stable populations of genome-edited hepatocytes. A second-generation PCSK9-specific meganuclease showed reduced off-target cleavage. These studies demonstrate efficient, physiologically relevant in vivo editing in non-human primates, and highlight safety considerations for clinical translation.

SUMMARYThe I-CreI homing endonuclease from Chlamydomonas reinhardti has been used as a molecular tool for creating DNA double-strand breaks and enhancing DNA recombination reactions in maize cells. The DNA-binding properties of this protein were re-designed to recognize a 22 bp target sequence in the 5th exon of MS26, a maize fertility gene. Three versions of a single-chain endonuclease, called Ems26, Ems26+ and Ems26++, cleaved their intended DNA site within the context of a reporter assay in a mammalian cell line. When the Ems26++ version was delivered to maize Black Mexican Sweet cells by Agrobacterium-mediated transformation, the cleavage resulted in mutations at a co-delivered extra-chromosomal ms26-site in up to 8.9% of the recovered clones. Delivery of the same version of Ems26 to immature embryos resulted in mutations at the predicted genomic ms26-site in 5.8% of transgenic T 0 plants. This targeted mutagenesis procedure yielded small deletions and insertions at the Ems26 target site consistent with products of doublestrand break repair generated by non-homologous end joining. One of 21 mutagenized T 0 plants carried two mutated alleles of the MS26 gene. As expected, the bi-allelic mutant T 0 plant and the T 1 progeny homozygous for the ms26 mutant alleles were male-sterile. This paper described the second maize chromosomal locus (liguless-1 being the first one) mutagenized by a re-designed I-CreI-based endonuclease, demonstrating the general utility of these molecules for targeted mutagenesis in plants.

Facile ''writing'' of DNA fragments that encode entire gene sequences potentially has widespread applications in biological analysis and engineering. Rapid writing of open reading frames (ORFs) for expressed proteins could transform protein engineering and production for protein design, synthetic biology, and structural analysis. Here we present a process, protein fabrication automation (PFA), which facilitates the rapid de novo construction of any desired ORF from oligonucleotides with low effort, high speed, and little human interaction. PFA comprises software for sequence design, data management, and the generation of instruction sets for liquid-handling robotics, a liquid-handling robot, a robust PCR scheme for gene assembly from synthetic oligonucleotides, and a genetic selection system to enrich correctly assembled full-length synthetic ORFs. The process is robust and scalable.