A peptide delivery system sneaks CRISPR into cells

The CRISPR-Cas9 system has developed into a powerful platform for genome editing in various types of cells and tissues with single-nucleotide precision, but limited delivery options hamper its application in real-world settings. A new study by Shen et al. describes the use of an amphipathic peptide to deliver Cas9/sgRNA ribonucleoprotein complexes, leading to the disruption of GFP genes in cells and mice. Disruption of the Nrip1 gene in isolated pre-adipocytes led to a “browning” phenotype, pointing to new options in the fight against diabetes and obesity.

Genome editing in mammalian cells, which introduces precise modifications to genomic DNA, holds great promise for treating human diseases. Compared with other DNA-targeted nucleases, such as zinc-finger nucleases (ZFNs) and transcription activator-like effector nucleases (TALEN), the clustered regularly interspaced short palindromic repeat (CRISPR) 2associated proteins 9 (CRISPR-Cas9) system has the advantage of easy programming of its target specificity, which has revolutionized the field of genome editing (1). In the CRISPR-Cas9 system, the Cas9 nuclease is recruited to a specific DNA locus by a single guide RNA (sgRNA), where the sgRNA binds to the target DNA site adjacent to a particular motif, such as a 5Ј-NGG-3Ј sequence (2,3). Cas9 then cleaves the DNA, introducing a double-strand break that will likely result in gene disruption during subsequent DNA repair.
Despite its potential, many challenges need to be addressed before translating this technology into the clinic (4,5). Among the numerous limitations, efficient delivery of the CRISPR-Cas9 system into the nuclei of cells without being entrapped in the endosome may be one of the most difficult challenges. Physical delivery approaches including microinjection, electroporation, and hydrodynamic injection have been shown to be efficient for transfection. But, they are not readily translatable for in vivo applications due to their low throughput and inappropriateness for human physiology (6). Recently, a variety of delivery vectors for CRISPR-Cas9 have been developed, including viral vectors, polymeric carriers, lipid or DNA nanoparticles, gold nanoparticles, and cell-penetrating peptides (7,8). However, most of these methods deliver nucleic acid forms of Cas9, which can cause delays in action due to transcription and/or translation of the protein or lead to concerns regarding specific dosing and off-target effects, especially in the case where a plasmid inserts into the genome, leading to sustained expression of the protein and sgRNA (6). Direct delivery of Cas9/sgRNA ribonucleoprotein could obviate these consequences, but an optimized delivery system for such a complex is lacking.
In this issue, Shen et al. (9) help to address this problem in their development of a new method termed CRISPR delivery particles (CriPs) for the delivery of CRISPR-Cas9 ribonucleoprotein complexes using an Endo-Porter (EP) peptide. Work from the Czech laboratory had previously demonstrated that the amphiphilic EP peptide could mediate delivery of siRNA particles, presumably facilitating intracellular transport by disrupting the endosomal membrane upon acidification, using its weak base residues to create a "proton-sponge effect." In the new study, the authors first demonstrated that the EP peptide could electrostatically interact with Cas9/sgRNA, coating the ribonucleoproteins into nanoparticles with a mean diameter of around 370 nm (Fig. 1a). The authors then confirmed the CriP complex is internalized and functional in a series of GFP-based model systems. Using a macrophage cell line and transgenic mice that stably express GFP, the efficacy of Gfp disruption in macrophage cells and primary cells, such as pre-adipocytes and peritoneal exudate cells (PECs), could be easily detected by measuring the intensity of green fluorescence. Specifically, they investigated the gene disruption efficacy of CriPs in a macrophage cell line and observed a 54% loss in Gfp gene (Fig. 1b), which showed higher efficiency than the control commercial transfection reagent Lipofectamine® RNAiMAX (30% GFP loss) in this study. They further examined the Gfp disruption efficacy in primary pre-adipocytes and PECs, where 48.5% and 40.7% GFP disruptions were observed, respectively. Furthermore, Gfp gene disruption was also demonstrated in vivo in a GFP mouse model, where systemically administered CriPs by i.p. injection lead to an average of 5.7% loss of GFP expression in PECs. Genomic changes were further confirmed by next-generation sequencing, where a 3% mutation rate was detected in contrast to the 0.02% background mutation.
Shen et al. (9) next investigated whether CriPs could disrupt disease-relevant genes. Considering the therapeutic potential of browning white adipocytes, which coverts high-fat "white fat cells" into low-fat "brown fat cells," in treating type 2 diabetes and obesity, they selected nuclear co-repressor Nrip1 as the target, which inhibits glucose utilization and fatty acid metab- The authors declare that they have no conflicts of interest with the contents of this article. 1 To whom correspondence should be addressed. E-mail: sunwj@ucla.edu. 2 The abbreviations used are: CRISPR, clustered regularly interspaced short palindromic repeats; CriP, CRISPR-delivery particle; sgRNA, single-guide RNA; EP peptide, Endo-Porter peptide; PECs, peritoneal exudate cells.
olism. To test this, primary white pre-adipocytes were isolated and treated with CriPs targeting Nrip1. After 8 days of differentiation, 43.8% Nrip1 mutation was observed, accompanied by an enhanced expression of the downstream uncoupling protein UCP1, suggesting the browning of white adipocytes (Fig. 1c). The innovative study by Shen et al. (9) reports an effective amphiphilic peptide-based nanocarrier, CriPs, for gene depletion in vitro and in vivo, with many advantages for therapeutic genome editing. First, delivery of Cas9/sgRNA as an active ribonucleoprotein allows for control over the dosage and timing; indeed, the authors did not detect any off-target editing in the differentiated adipocyte subjects to Nrp1 disruption. Second, the carrier is simply a peptide, which should be intrinsically biocompatible and degradable. Third, the preparation of CriPs is very simple, without sophisticated chemical modification steps, making it easier for this approach to be translated to other studies and to the clinic. Last but not least, successful gene depletion in vivo achieved by the CriPs system demonstrated the potential of this approach for direct editing of somatic cells, where it can be programmed to treat human diseases if a therapeutic gene is targeted (10).
Overall, CriPs showed reasonably high in vitro gene-editing efficacy, which could be a powerful tool in generating therapeutic cells, for example in knocking out the checkpoint receptors of immune cells. Combined with downstream purification of edited cells, 50% gene disruption efficiency could be very helpful in producing therapeutic cells. As Shen et al. (9) did not explore the disruption of Nrip1 in vivo, we suspect that the 5% success rate in the GFP model system is not yet sufficient to gain therapeutic benefits in the complicated metabolic biology. However, the level of gene editing necessary is not clear. For example, it has been reported that 40% disruption of Pcsk9 in vivo lead to substantial reduction in blood cholesterol level, and 1% efficacy in the repair of factor IX could bring about health benefits in hemophilia B treatment. In any case, further improvement of in vivo gene-editing efficacy of the peptide carrier, such as optimizing the sequence of the peptide or incorporating other types of materials, would undoubtedly expand its applicability to more diseases. After conquering the challenge of delivery efficacy, potential side effects, like unintended accumulation in healthy organs or off-target editing of other genomic locations, needed to be investigated systemically for regulatory approvals. Endo-Porter (EP) peptide-mediated delivery of Cas9/sgRNA for the reporter and therapeutic gene disruption. a, the CriPs system was prepared through electrostatic interaction between Cas9/sgRNA ribonucleoprotein and EP peptide. b, Gfp gene depletion was tested in reporter macrophage cell line, pre-adipocyte, and PECs. c, disruption of the Nrp1 gene leads to a browning phenotype of white adipocyte.